This version of the article has been accepted for publication, after peer review (when applicable) and

is subject to
Springer Nature’s AM terms of use, but is not the Version of Record and does not reflect post-acceptance improvements,
or any corrections. The Version of Record is available online at: http://dx.doi.org/10.1007/s11069-020-04124-3
Noname manuscript No.
(will be inserted by the editor)

1 Applications of artificial intelligence for disaster

2 management

3 Wenjuan Sun · Paolo Bocchini ·

4 Brian D. Davison

6 the date of receipt and acceptance should be inserted later

7 Abstract Natural hazards have the potential to cause catastrophic damage

8 and significant socioeconomic loss. The actual damage and loss observed in the
9 recent decades has shown an increasing trend. As a result, disaster managers
10 need to take a growing responsibility to proactively protect their communities
11 by developing efficient management strategies. A number of research studies
12 apply artificial intelligence (AI) techniques to process disaster-related data for
13 supporting informed disaster management. This study provides an overview of
14 current applications of AI in disaster management during its four phases: miti-
15 gation, preparedness, response, and recovery. It presents example applications
16 of different AI techniques and their benefits for supporting disaster manage-
17 ment at different phases, as well as some practical AI-based decision support
18 tools. We find that the majority of AI applications focus on the disaster re-
19 sponse phase. This study also identifies challenges to inspire the professional
20 community to advance AI techniques for addressing them in future research.

21 Keywords Disaster resilience · Disaster management · Artificial intelligence

Wenjuan Sun
Department of Civil and Environmental Engineering, Lehigh University, Bethlehem, PA,
United States, 18015.
E-mail: wes316@lehigh.edu
Paolo Bocchini
Department of Civil and Environmental Engineering, Lehigh University, Bethlehem, PA,
United States, 18015.
E-mail: paolo.bocchini@lehigh.edu
Brian D. Davison
Department of Computer Science and Engineering, Lehigh University, Bethlehem, PA,
United States, 18015.
E-mail: davison@cse.lehigh.edu
 2 Wenjuan Sun et al.

22 1 Introduction

23 Natural hazards have caused catastrophic damage and significant socioeco-

24 nomic loss, showing an increasing trend (Hoeppe 2016). Statistics for 2017
25 indicate economic losses from natural hazards in the United States exceed
26 $300 billion; Hurricane Harvey alone has caused $125 billion in socioeconomic
27 losses (Wilts 2018). These adverse impacts pose challenges to disaster response
28 managers, who face increasingly tight resources and an exhausted workforce,
29 and such challenges force local authorities to re-evaluate their policies for dis-
30 aster management.
31 There are large volumes of data generated daily, including real data and
32 simulation data. Both types of data can be used to support disaster manage-
33 ment. The advancement of information communication technologies, such as
34 social media, telecommunication data, and remote sensing, make large volumes
35 of real data available (Eguchi et al. 2008; Boccardo and Tonolo 2014; Rawat
36 et al. 2015; Adeel et al. 2018; Novellino et al. 2018). Sometimes, real data is
37 scarce. In research communities, many computational models are developed
38 to generate simulation data for estimating the disaster-induced impact and
39 identifying vulnerable structures, such as IN-CORE (Ellingwood et al. 2016)
40 and PRAISys (The PRAISys Team 2018). Regardless of data type, acquir-
41 ing, managing, and processing big data in a short time is essential to support
42 efficient disaster management. Using AI to analyze the voluminous data to
43 rapidly extract useful and reliable information becomes increasingly popular
44 for supporting effective decision-making in disaster management (Eskandar-
45 pour and Khodaei 2017; Velev et al. 2018; Yu et al. 2018; Wang et al. 2018d;
46 Barabadi and Ayele 2018).
47 Some published studies have reviewed AI applications in disaster man-
48 agement, with the topic targeted to certain types of hazard, infrastructure,
49 and data. For example, Fotovatikhah et al. (2018) have discussed the status
50 and challenges of applying computational intelligence methods to major flood
51 control and disaster management. Zagorecki et al. (2013) have reviewed ap-
52 plications of data mining and machine learning to disaster management, but
53 there is no discussion on any practical AI-based decision support tools. Other
54 studies review how computer vision methods have been applied for disaster
55 management by analyzing remote sensing data, such as target recognition with
56 deep learning (Zhang et al. 2016b), fire detection with wavelet analysis and
57 neural networks (Yuan et al. 2015), and estimating three-dimensional struc-
58 tures (Gomez and Purdie 2016). However, very few of them have explicitly
59 discussed the progress and challenges of how AI has been applied in disaster
60 management in different phases, by considering hazard and infrastructure as
61 well as data in a general sense.
62 In what follows, we describe the research background of AI methods and
63 disaster management first, followed by the state of research and practice of
64 applications of AI in disaster management in four phases, and the challenges
65 therein. In particular, practical decision support tools for disaster manage-
66 ment based on AI methods have been reviewed. This study can facilitate new
 Applications of artificial intelligence for disaster management 3

67 researchers to identify critical research gaps in this field and provide practi-
68 tioners a comprehensive summary for selecting an appropriate AI model and
69 practical decision support tool based on their community needs.

70 2 Background

71 2.1 AI methods

72 This study reviews the state of research and practice of applying AI in dis-
73 aster management, by classifying AI methods in six categories: supervised
74 models, unsupervised models, deep learning, reinforcement learning, and deep
75 reinforcement learning, as well as optimization.

76 2.1.1 Supervised models

77 Supervised models represent algorithms that are trained on pre-existing data

78 with human input. Using labelled training data with known input and out-
79 put pairs, supervised models infer a function from input to output using re-
80 gression/classification methods to predict the value/category of the output
81 variable (Russell and Norvig 2016). In general, supervised models have been
82 used for information extraction, object recognition in computer vision, pattern
83 recognition, and speech recognition, etc.

84 2.1.2 Unsupervised models

85 Without human input, unsupervised models use statistical methods to extract

86 hidden structure in unlabeled data based on inherent characteristics (Rus-
87 sell and Norvig 2016). Unsupervised models are suitable for detecting the
88 abnormal data and reducing the data dimension, with wide applications to
89 clustering and data aggregation problems. Clustering algorithms are used for
90 pattern recognition by partitioning unlabeled data into multiple groups based
91 on certain similarity features (Maulik and Bandyopadhyay 2002). Dimension
92 reduction algorithms, such as principal component analysis (PCA), can reduce
93 the complexity of data and avoid overfitting.

94 2.1.3 Deep learning

95 Deep learning is a class of algorithms that use multiple layers to extract fea-
96 tures from the input data progressively, with improved learning performance
97 and broad application scopes (Deng and Yu 2014; Pouyanfar et al. 2018). De-
98 spite the drawback of requiring long training time, deep learning algorithms
99 are particularly suitable to solve problems of damage assessment, motion de-
100 tection, and facial recognition, transportation prediction, and natural language
101 processing for supporting disaster management. For example, recursive neural
102 networks (RvNN) and recurrent neural networks (RNN) have been successfully
 4 Wenjuan Sun et al.

103 applied to natural language processing (NPL) (Socher et al. 2011; Graves et al.
104 2013). Convolutional neural networks (CNN) are suitable for image recogni-
105 tion (Simonyan and Zisserman 2014), computer vision (Krizhevsky et al. 2017),
106 NPL (Zhao and Wu 2016), and speech processing (Dahl et al. 2012).

107 2.1.4 Reinforcement learning

108 By learning from a series of reinforcements (using punishment and rewards as

109 positive and negative signals), reinforcement learning algorithms are modeled
110 in the form of Markov decision processes to address goal-oriented problems
111 for making decisions in a sequential manner (Russell and Norvig 2016). Re-
112 inforcement learning is suitable for solving problems that need to make a se-
113 quence of decisions in an uncertain and complex environment, with successful
114 applications in robotics, resource management, and traffic light control. The
115 main challenge in reinforcement learning is preparing the suitable training
116 environment that is closely related to tasks to be performed. Typical rein-
117 forcement learning algorithms include Q-learning and SARSA (State-Action-
118 Reward-State-Action), to name a few (Sutton and Barto 2018).

119 2.1.5 Deep reinforcement learning

120 Deep reinforcement learning combines reinforcement learning with deep neural
121 networks with the aim of creating software agents that can learn by themselves
122 to establish successful policies for gaining the most long-term rewards. Deep
123 reinforcement learning has superior performance for solving problems with
124 complex sequential tasks, such as computer vision, robotics, finance, smart
125 grids, etc. Requiring a large amount of training data and training time to
126 reach reasonable performance, deep reinforcement learning sometimes becomes
127 extremely computationally expensive.

128 2.1.6 Optimization

129 While the focus of this study is how AI methods are applied for disaster
130 management, optimization is an essential ingredient in most of AI methods to
131 find the best model as measured by an objective function. For this reason, this
132 study explicitly lists three optimization techniques as example methods and
133 investigates their applications in disaster management.

134 2.2 Disaster management

135 2.2.1 Four phases of disaster management

136 As shown in Fig. 1, disaster management involves four phases: mitigation, pre-
137 paredness, response, and recovery. The mitigation phase refers to management
138 activities for preventing or minimizing future emergencies and consequences
 Applications of artificial intelligence for disaster management 5

Before a Disaster During a Disaster After a Disaster

Prepared- Recovery
Mitigation Response
ness

Develop Stock disaster Search and Debris removal

preventive laws supplies kit rescue to Precise damage
and regulations Develop mutual identify affected assessment
Implement aid agreements people Infrastructure
advanced codes and plans Assess initial destruction and
and standards Train response damage reconstruction
Establish zoning personnel and Provide first-aid Restore the
requirements concerned and livelihoods
Buy insurance citizens humanitarian Community
Construct Prepare assistance development
barriers shelters and Open and
backup manage shelters
facilities

Fig. 1 Four phases of disaster management.

139 with long-term benefits. Examples of mitigation activities include enforcing

140 advanced building codes and standards, retrofitting highway overpasses, hos-
141 pitals, and shelters, informing and educating the general public and related
142 stakeholders about hazards and potential mitigation strategies. The prepared-
143 ness phase comes into place when an emergency or a disaster is likely to take
144 place. It corresponds to preparatory activities prior to a disaster in order to
145 save lives and help response and rescue operations, such as stocking food and
146 water, posting emergency contacts, and preparing evacuations. With plans and
147 strategies developed beforehand, the response phase mainly puts them into
148 action. Response activities happen during a disaster, usually involving evacu-
149 ating threatened areas, firefighting, search and rescue efforts, shelter manage-
150 ment, and humanitarian assistance. After a disaster, the recovery phase refers
151 to repair and reconstruction efforts to return to a normal or even better func-
152 tionality level. Recovery actions usually include debris cleanup, precise damage
153 assessment, and infrastructure reconstruction, as well as financial assistance
154 from government agencies and insurance companies.

155 2.2.2 Disaster management and disaster resilience

156 The goals of disaster management are to implement operations and strategies
157 to effectively prepare, rapidly respond and rescue, efficiently allocate resources,
158 quickly correct damage and recover to full functionality, ultimately protect the
159 community and minimize the adverse impact. That is to say that the efficient
160 disaster management should strengthen the disaster resilience of a community.
161 The term “disaster resilience” refers to the ability of an entity to anticipate,
162 resist, absorb, adapt to, and rapidly recover from an unexpected disturbance
163 (DHS 2010). Fig. 2 displays features of disaster resilience in terms of dimen-
164 sions, stakeholders, disruption types, properties of resilient entities, and ben-
165 efits. In case of a disaster, such as a hurricane or an earthquake, a resilient
 6 Wenjuan Sun et al.

Technical
Organizational
Economic
Social
Health

Disaster
resilience

Fig. 2 Features of disaster resilience.

166 community is expected to be able to protect people, infrastructure, and socioe-

167 conomic environment, with reliable performance and fast recovery capability,
168 as well as minimal adverse consequence. The disaster resilience of a community
169 can be enhanced by improving the rapidity, robustness, resourcefulness, and
170 redundancy, as well as learning capability, in which learning refers to residents’
171 changing expectations with respect to infrastructure performance and opera-
172 tional adaptations of infrastructures to new circumstances during and after a
173 disaster (Sun et al. 2020b). From the disaster management perspective, gov-
174 ernments and other stakeholders organize their operations in multiple aspects
175 (technical, organizational, economic, social, and health), various management
176 plans and strategies are developed and implemented.
177 A number of programs have been established to promote the research and
178 practice of disaster resilience for supporting informed decision-making in dis-
179 aster management. Some examples in the United States are described as fol-
180 lows. Since 2013, the Campus Resilience Program has yielded successful tools
181 and guidelines for evaluating the vulnerability of the academic community
182 nationwide. The Hazard Mitigation Grant Program (HMGP) supports com-
183 munities in implementing cost-effective hazard mitigation measures, such as
184 structure retrofit and reconstruction, to eliminate the risk of loss of life and
185 property damage from future disasters (FEMA 2018). The Community Re-
186 silience Planning Guide presents a six-step process to help local community
187 authorities identify gaps, create resilience plans, and implement strategies for
188 better community resilience against future disasters (NIST 2018; Cauffman
 Applications of artificial intelligence for disaster management 7

189 et al. 2018). In addition, local authorities and private organizations have been
190 implementing practices for resilience enhancement. For example, Los Angeles
191 County in California has developed a community resilience toolkit to support
192 decision-making in disaster management (Eisenman et al. 2014; Bromley et al.
193 2017). The 100 Resilient Cities program supports city governments’ efforts in
194 fostering urban resilience and addressing climate change and equity (The Rock-
195 efeller Foundation 2019). In parallel, other countries have also been actively
196 working in this direction. The Horizon 2020 Research and Innovation Pro-
197 gramme has developed the European Resilience Management Guideline and
198 tools for supporting effective disaster management and enhancing the resilience
199 against disasters and climate change (EU-CIRCLE 2019). Under the Sendai
200 Framework for Disaster Resilience Network, the Asia-Pacific region has been
201 undertaking major reforms in developing disaster management policies with
202 increasing applications of AI in disaster response (UN 2015; Renwick 2017;
203 Pau et al. 2017; Izumi et al. 2019). All these guidelines and computational
204 tools aim to support disaster management and enhance disaster resilience. AI
205 has great potential to alleviate the burden of decision makers in disaster man-
206 agement by processing large amounts of disaster-related data more efficiently
207 and effectively.

208 3 Applications of AI for Disaster Management

209 Fig. 3 shows the increasing trend in the number of publications on World-
210 Cat from 1991 to 2018 with regards to applying AI to disaster management.
211 The greatest number of publication in disaster response among four phases
212 indicates that applications of AI mainly focus on this phase. While AI will
213 not replace the experience and wisdom of well-trained disaster professionals,
214 at least in the foreseeable future, AI techniques can rapidly analyze big data
215 and perform predictive analytics for supporting decision-making in disaster
216 management.
217 To illustrate how different AI methods have been applied in disaster man-
218 agement, we have identified a total of 26 AI methods and 17 application areas
219 as representative examples. By using every AI method and every application
220 area as key words, we have searched for related literature on the websites
221 of Google Scholar and Web of Science, requiring joint presence of both key-
222 words. Figure 4 presents our findings on AI applications to the four phases
223 and their sub-areas. In this figure, every solid line demonstrates the presence
224 of applications of an AI method in a certain area. More solid lines connecting
225 to Application Areas 1 ∼ 4 and 9 ∼ 13 mean that there are more studies
226 applying AI methods in mitigation and response phases. Detailed application
227 examples are presented as citations in Tables 1 ∼ 4. It is worth noting that
228 only the most relevant/representative publications are presented in some cells
229 in the tables due to space limits.
 8 Wenjuan Sun et al.

3000
Disaster management & artificial intelligence
2500 Disaster mitigation & artificial intelligence
Disaster preparedness & artificial intelligence
No. of publications

Disaster response & artificial intelligence

2000 Disaster recovery & artificial intelligence

1500

1000

500

0
4 8 2 6 0 4 8
99 99 00 00 01 01 01
9 1-1 95-1 99-2 0 3-2 0 7-2 11-2 15-2
19 19 19 20 20 20 20
Year
Fig. 3 An increasing number of publications on artificial intelligence in disaster manage-
ment.
Note: Publications refer to articles, books, and downloadable archive materials. The num-
ber of publications is determined by summing the number of publications every four years
between 1991 and 2018 when searching with the keywords in the legend on WorldCat
(http://www.worldcat.org/).

230 3.1 AI Applications in Disaster Mitigation

231 In the disaster mitigation phase, decision makers need to identify hazard and
232 risks (Application Area 1), predict possible impact (Application Area 2), assess
233 vulnerability (Application Area 3), and develop mitigation strategies (Applica-
234 tion Area 4), in order to create stronger, safer, and more resilient communities.
235 AI methods have been widely applied to support disaster mitigation manage-
236 ment in the four areas. In particular, supervised models and unsupervised
237 models have been extensively used for Application Area 1, followed by Areas
238 2 and 3. Conversely, reinforcement learning and deep reinforcement learning
239 are rarely used in the four areas.
240 Possible hazards and risks should be identified for the community of in-
241 terest. For natural hazards, characteristics of terrain, lithology, meteorology,
242 and even human activities should be analyzed, and hazard zone maps should
243 be developed. Traditional methods, such as field monitoring, physics-based
244 models, expert surveys, and multi-criteria decision-making methods, are ap-
245 plied to identify hazards and risk factors. Sometimes, these methods are labor
246 intensive, possibly with high false alarm rate (Bellaire et al. 2017). In this
247 case, AI techniques can rapidly analyze large volumes of data to assess hazard
248 risks in a timely manner (Pradhan 2009; Yilmaz 2010). There are extensive
249 studies applying different AI methods to developing susceptibility maps for
250 different types of hazards. For instance, snow avalanche predictions have been
251 made using logistic regression (LR) (Gauthier et al. 2017), support vector
 Applications of artificial intelligence for disaster management 9

Artificial intelligence method Application area Phase

A Linear regression
Forecast hazard and assess
1
B Non-linear regression risk

C Logistic regression Evaluate possible impacts

Mitigation
2
due to hazards
Supervised models

D Support vector machine

3 Assess vulnerability
E Naïve Bayes
Develop/Compare
4
F Decision tree mitigation strategies

G Random forest Early warning/alert

5
systems
H K-nearest neighbors

Preparedness
Real-time disaster
6
I Logistic model tree prediction and detection

J Neural networks 7 Training systems

K Hierarchical clustering
Unsupervised models

8 Disaster evacuation
L K-means clustering

M Fuzzy clustering 9 Event mapping

N Principal component analysis

10 Damage assessment
O Hidden Markov models

Response
Disaster rescue and relief,
11
P Convolutional neural networks resource allocation
Deep learning

Q Recurrent neural networks Disaster information system

12
and inter-agency collaboration
R Deep neural network
Understand people’s concern,
13
emotion, and reaction
S Multi-layer perceptron

T Recursive neural network 14 Assess impact

U Q-learning
Develop post-event
RL

Recovery
15
V Policy gradient recovery plan
Optimization DRL

W Deep Q-networks 16 Track recovery

X Genetic algorithm Evaluate loss and repair

17
cost
Y Particle swarm optimization
RL = Reinforcement learning
Z Simulated annealing
DRL = Deep reinforcement learning

Fig. 4 Applications of artificial intelligence in disaster management.

Note: A solid link between an AI method and an application area represents the fact that there
are applications of the AI method to this area. Detailed application examples are presented in
Tables 1 ∼ 4.

252 machine (SVM) (Choubin et al. 2019), and neural networks (Dekanová et al.
253 2018; Rauter and Winkler 2018). Landslide susceptibility can be assessed by
254 SVM (Xu et al. 2012; Goetz et al. 2015; Zhou et al. 2018a), LR (Goetz et al.
255 2015; Zhou et al. 2018a), random forest (RF) (Goetz et al. 2015), and neural
256 networks (Dou et al. 2015; Zhou et al. 2018a). The aforementioned AI meth-
257 ods have also been applied to other types of hazards, such as mapping forest
258 fire susceptibility (Sachdeva et al. 2018), predicting fire size (Mitsopoulos and
259 Mallinis 2017), and forecasting precipitation (Huang et al. 2018).
260 AI techniques have been applied to estimate possible impacts and assess
261 vulnerability. For instance, possible structural damage under natural hazard(s)
262 can be predicted by using fragility curves, which were traditionally built from
 10 Wenjuan Sun et al.

263 statistical analyses of historical and simulation data and now can be estimated
264 from the application of AI methods, such as LR (Ghosh et al. 2013; Kamesh-
265 war and Padgett 2014; Mangalahtu et al. 2018), neural networks (Lagaros
266 and Fragiadakis 2007; Mitropoulou and Papadrakakis 2011; Liu and Zhang
267 2018; Mangalathu et al. 2018), and SVM (Mahmoudi and Chouinard 2016).
268 Infrastructure service disruptions due to hazards can be predicted based on his-
269 torical data using generalized regression models (Reed 2008; Liu et al. 2008),
270 RF (Nateghi et al. 2014; Cerrai et al. 2019; D’Amico et al. 2019), decision
271 tree (DT) (Wanik et al. 2015), and Bayesian additive regression tree (BART)
272 (Cerrai et al. 2019). Using data from physical sensors and social sensing, the
273 vulnerability of structures and communities can be assessed with spatial regres-
274 sion models (Wang et al. 2019g), RF (Yoon and Jeong 2016), neural networks
275 (Wu et al. 2008), deep neural networks (Nabian and Meidani 2018b), etc. In
276 terms of the number of publications, there are fewer applications of AI methods
277 to estimating hazard-induced impact and assessing community vulnerability
278 (Application Areas 2 and 3), compared with those on hazard forecast and risk
279 assessment (Application Area 1).
280 Based on the impact and vulnerability analyses, decision makers can gain
281 better situation awareness with more confidence and develop effective miti-
282 gation strategies (Schwartz 2018), such as retrofitting vulnerable structures
283 (Karamlou et al. 2016), elevating electric substations and using underground
284 cables (Duffey 2019), and developing effective disaster-related policies (Sun
285 et al. 2020a, 2021). In this process, AI techniques can support developing and
286 comparing mitigation strategies. For instance, different AI methods have been
287 applied to identifying management priorities (Canon et al. 2018), estimat-
288 ing people’s needs during a disaster (Nguyen et al. 2019a), and recognizing
289 human activities (Sadiq et al. 2018). Clustering algorithms are used for an-
290 alyzing remote images and developing contingency plans (Dou et al. 2014),
291 and optimization algorithms have been applied for developing effective plans
292 of disaster response and restoration (Bocchini and Frangopol 2012a,b; Gama
293 et al. 2016). So far, there are only a very small number of studies that apply
294 AI to developing and comparing mitigation strategies (Application Area 4).

295 3.2 AI Applications in Disaster Preparedness

296 In the preparedness phase, decision-makers should send out early warnings
297 and alert the public (Application Area 5) after identifying the disaster that is
298 about to come (Application Area 6), utilize emergency training systems and
299 tools (Application Area 7), and prepare for evacuations if needed (Application
300 Area 8). Among the four areas, most AI methods have been applied to Areas
301 5, 6, and 8, with very limited applications to Area 7.
302 Identifying the coming disasters in real time and sending out early warnings
303 are practical solutions for disaster preparations. These tasks usually rely on
304 experts’ analyses and judgments of sensor measurements in the field, and AI
305 techniques can serve as an alternative in a cost-effective manner to forecasting
Applications of artificial intelligence for disaster management 11

Table 1 Example AI applications for disaster mitigation

AI Method 1. Forecast hazard and 2. Estimate impact 3. Assess vulnerability 4. Develop/Compare
risk strategy
A. Linear regression & Reed (2008); Chang Kahn (2006); Simmons Yang and Yu (2011); NA
extensions et al. (2010); Kim and Sutter (2008); Geiß et al. (2014); Heß
et al. (2019) Zahran et al. (2008); (2017); Wang et al.
Peduzzi et al. (2009); (2019g); Sun et al.
Maliszewski et al. (2019)
(2012)
B. Non-linear Pradhan (2009); Zorn and Shamsedin NA NA
regression Yilmaz (2010); (2015); Lee et al.
Trafalis et al. (2014); (2016)
Lin et al. (2017a);
Goetz et al. (2015)
C. Logistic regression Bai et al. (2010); Eskandarpour and Ettinger et al. (2016); Khan and Sayem
Marjanović et al. Khodaei (2017); Li et al. (2019b) (2012); Rakgase and
(2011); Ozdemir and Rosellini et al. (2018); Norris (2014);
Altural (2013); Wang Yuan and Moayedi Cavalcante et al.
et al. (2013b) (2019) (2019)
D. Support vector Yilmaz (2010); Galatzer-Levy et al. Geiß et al. (2014); Sun Guo et al. (2009);
machine Marjanović et al. (2014); Li et al. (2014); et al. (2019); Xiong Rudin et al. (2012);
(2011); Xu et al. Karstoft et al. (2015); et al. (2019) Dou et al. (2014);
(2012); Lin et al. Tinoco et al. (2018) Pogrebnykov and
(2017a); Zhou et al. Maldonado (2017)
(2018a)
E. Naı̈ve Bayes Shirzadi et al. (2017); Bawono et al. (2020) Geiß et al. (2014) Sadiq et al. (2018)
Chen et al. (2019);
Sankaranarayanan
et al. (2019)
F. Decision tree Saito et al. (2009); Wanik et al. (2015); Sriram et al. (2019) Guo et al. (2009);
Marjanović et al. Yuan and Moayedi Sadiq et al. (2015,
(2011); Rhee and Im (2019) 2018)
(2017)
G. Random forest McGovern et al. Galatzer-Levy et al. Yoon and Jeong Rudin et al. (2012)
(2011); Goetz et al. (2014); Nateghi et al. (2016); Sriram et al.
(2015); Rhee and Im (2014); Wanik et al. (2019)
(2017); Chen et al. (2015); Cerrai et al.
(2018) (2019)
H. K-nearest neighbors Liu et al. (2016); Cheng and Hoang Leon and Atanasiu Sun et al. (2017); Sadiq
Sankaranarayanan (2014) (2006); Kusumawardani et al. (2015, 2018)
et al. (2019) et al. (2016)
I. Logistic model tree Chen et al. (2018, NA Yang et al. (2019d) NA
2019)
J. Neural networks Melchiorre et al. Karamouz et al. Wu et al. (2008); Jones et al. (2008)
(2008); Yilmaz (2010); (2014); Tinoco et al. Pilkington and
Dou et al. (2015); (2018); Oktarina et al. Mahmoud (2016); Guo
Huang et al. (2018) (2019); Tinoco et al. et al. (2018); Wahab
(2019) and Ludin (2018)
K. Hierarchical Leśniak and Isakow NA Cavalieri et al. (2014); NA
clustering (2009); Trugman and Su et al. (2015); Kim
Shearer (2017) et al. (2017); Chang
et al. (2018)
L. K-means clustering Iliadis (2005); Lam et al. (2016) Su et al. (2015); Pual2012 (2012)
Melchiorre et al. Fernandez et al. (2016)
(2008); Leśniak and
Isakow (2009);
Jayaram and Baker
(2010)
M. Fuzzy clustering Zhang (2004); Shi da Silva et al. (2008); Alam et al. (2000); Wu Dou et al. (2014)
et al. (2010); Wang Wlwood and Corotis et al. (2013); Chen
et al. (2013b); Ansari (2015) et al. (2014b)
et al. (2015); Wang
et al. (2018c)
N. Principle Chen and Hong Li et al. (2014) Chen et al. (2014a); Moradi et al. (2019)
component analysis (2012); Shi et al. Fernandez et al.
(2015) (2016); Heß (2017);
Uddin et al. (2019)
O. Hidden Markov Wang et al. (2010b); Song et al. (2014, NA Eicken et al. (2011)
models Khadr (2016); Wang 2016)
et al. (2018a)
P. Convolutional DeVries et al. (2018); NA Crawford et al. (2018); Pogrebnykov and
neural networks Padmawar et al. Han et al. (2019) Maldonado (2017);
(2019) Nguyen et al. (2019a)
Q. Recurrent neural Ma et al. (2015b); NA NA Canon et al. (2018);
networks Asim et al. (2017); Pechenkin and
Cortez et al. (2018); Demidov (2018);
Wang et al. (2020b); Nguyen et al. (2019a);
Mutlu et al. (2019) Yang et al. (2019b)
R. Deep neural Sankaranarayanan NA Nabian and Meidani NA
network et al. (2019) (2018b); Dogaru and
Dumitrache (2019)
S. Multi-layer Zare et al. (2013); Yuan and Moayedi Wahab and Ludin Sadiq et al. (2018)
perception Hernández et al. (2019) (2018)
(2016); Pham et al.
(2017)
T. Recursive neural Mishra and Desai NA NA NA
network (2006);
Hosseini-Moghari and
Araghinejad (2015)
U. Q-learning Lin et al. (2013) NA Yan et al. (2016); Zhang et al. (2019b)
Otoum et al. (2019)
V. Policy gradient NA NA NA NA
W. Deep Q-networks NA NA NA Elsayed and
Erol-Kantarc (2018)
X. Genetic algorithm Chang and Chien Tinoco et al. (2019) NA Tapia and Padgett
(2007); Terranova (2015); Yan et al.
et al. (2015) (2017); Yang et al.
(2019b)
Y. Particle swarm Romlay et al. (2016); NA NA NA
optimization Padmawar et al.
(2019)
Z. Simulated annealing Zhu and Wu (2013); NA NA Afandizadeh et al.
Hosseini et al. (2019) (2013); Ma et al.
(2015a); Gama et al.
(2016)
NA = no literature was found on the application area (column) using the AI method (row).
12 Wenjuan Sun et al.

Table 2 Example AI applications for disaster preparedness

AI Method 5. Early warning 6. Real-time disaster 7. Training systems 8. Disaster evacuation
system prediction and detection
A. Linear regression Uunk et al. (2010); NA NA NA
Nolasco-Javier and
Kumar (2018); Pillai
et al. (2019)
B. Non-linear Moon et al. (2018) NA NA NA
regression
C. Logistic regression Wang et al. (2013a); Agarwal et al. (2016); NA Riad et al. (2006);
Hoot and Aronsky Kong et al. (2016b); Nguyen et al. (2016)
(2006) Zhao et al. (2020)
D. Support vector Sakaki et al. (2012); Arridha et al. (2017); de NA Mori et al. (2013);
machine Chou and Thedja Morsier et al. (2013); Higuchi et al. (2014);
(2016); Rafiei and Grasic et al. (2018); Jiang et al. (2017);
Adeli (2017); Wang Jhong et al. (2017); Wang et al. (2019b)
et al. (2019c); Mori Zhao et al. (2020)
et al. (2013);
Pogrebnykov and
Maldonado (2017)
E. Naı̈ve Bayes Mane and Mokashi Muda et al. (2011); NA NA
(2015) Kumar et al. (2014);
Grasic et al. (2018)
F. Decision tree Chen and Wang Arridha et al. (2017) NA Burris et al. (2015);
(2009); Zhou et al. Wang et al. (2019b)
(2017a)
G. Random forest Li et al. (2018b); Grasic et al. (2018); Yu NA NA
Moon et al. (2018) et al. (2017)
H. K-nearest neighbors Pyayt et al. (2011); Muda et al. (2011); NA Rahman and Hasan
Cheng et al. (2013); Kumar et al. (2014) (2018); Wang et al.
Ali et al. (2019); (2019b)
Tomin et al. (2013)
I. Logistic model tree NA NA NA NA
J. Neural networks Duncan et al. (2013); Ren et al. (2010); Bande Djordjevich et al. Sharma and Ogunlana
Kong et al. (2016a); and Shete (2017); (2008) (2015); Nguyen et al.
Moon et al. (2018); Berkhahn et al. (2019); (2016); Rahman and
Muhammad et al. Zhao et al. (2020) Hasan (2018); Peng
(2018); Abdullahi et al. (2019); Wang
et al. (2018); Tomin et al. (2019b)
et al. (2013)
K. Hierarchical NA Ifrim et al. (2014); NA Özdamar and Demir
clustering Akhtar and Siddique (2012)
(2017)
L. K-means clustering Naidu et al. (2018); NA NA Andersson et al. (2012)
Tomin et al. (2013)
M. Fuzzy clustering Saad et al. (2014); Ren et al. (2010) NA NA
Tomin et al. (2013)
N. Principal Peiris et al. (2010); NA NA NA
component analysis Wan and Mita (2010)
O. Hidden Markov Holgado et al. (2017) Benı́tez et al. (2007); NA Andersson et al.
models Toreyin and Cetin (2012); Raymond et al.
(2009); Günay et al. (2012); Song et al.
(2010); Heck et al. (2015)
(2010)
P. Convolutional Cheng et al. (2017); Ali et al. (2019); Layek NA NA
neural networks Lohumi and Roy et al. (2019); Wang
(2018); Perol et al. et al. (2019a);
(2018); Long et al. Muhammad et al.
(2018); Giffard-Roisin (2018)
et al. (2018);
Muhammad et al.
(2018); Pogrebnykov
and Maldonado (2017)
Q. Recurrent neural Hoot and Aronsky Chen et al. (2013); NA Rahman and Hasan
networks (2006); Cheng et al. Chang et al. (2014); (2018)
(2017); Pogrebnykov Jaech et al. (2019)
and Maldonado
(2017); Long et al.
(2018)
R. Deep neural Long et al. (2018) NA NA Jiang et al. (2017)
network
S. Multi-layer Khan et al. (2018) Tian and Chen (2017a); NA NA
perception Wang et al. (2019a)
T. Recursive neural NA NA NA NA
network
U. Q-learning NA Lingam et al. (2019) Khouj et al. (2011) Sarabakha and
Kayacan (2016); Yao
et al. (2019)
V. Policy gradient NA NA NA Zheng and Liu (2019)
W. Deep Q-networks NA NA NA Sharma et al. (2020)
X. Genetic algorithm Shirzaei and Walter Ahmad et al. (2009) NA Pourrahmani et al.
(2010); Terranova (2015); Sharma and
et al. (2015) Ogunlana (2015); Gao
et al. (2019)
Y. Particle swarm Huang and Xiang Lingam et al. (2019) NA Wang et al. (2010a);
optimization (2018) Zheng et al. (2013b)
Z. Simulated annealing NA Zhang et al. (2016a) NA Jahangiri et al. (2011)
 Applications of artificial intelligence for disaster management 13

306 the coming events (Ko and Kwak 2012), such as impending hurricane tra-
307 jectories and storms (Ghosh and Krishnamurti 2018), earthquakes (Mousavi
308 et al. 2019), ice jams (Zhao et al. 2012), floods (Yaseen et al. 2015), volcano
309 eruptions (Parra et al. 2016), and fires (Muhammad et al. 2018). For instance,
310 the UrbanFlood project in Europe has established an internet-based platform
311 for early flood warnings, in which an AI component has been developed for
312 detecting abnormal dike behaviours based on the analysis of thousands of sen-
313 sor streams (Noymanee et al. 2017). Sakaki et al. (2012) performed semantic
314 analysis of Japanese tweets with a tweet crawler, estimated the earthquake
315 location, and developed a reporting system named Toretter that was faster
316 than broadcast announcements by Japan Meteorological Agency. Based on
317 the real-time analysis of smartphone accelerometer measurements of tilting
318 motions, earthquake early warnings can also be sent out (Reilly et al. 2013).
319 Prior to a disaster event, utility companies can use AI-based tools to estimate
320 likely damage locations and service outage duration and get prepared before-
321 hand. For example, Hydro One, a large utility company in Ontario, Canada,
322 has successfully used such real-time predictive analyses in April 2018 and then
323 positioned crews in key areas and effectively restored the power service within
324 four days, significantly reducing the restoration time (McConnon 2018). With
325 the implementation of IoT, cloud network services can also rapidly and accu-
326 rately share information on disaster situations for early warnings (Chung and
327 Park 2016).
328 With respect to disaster evacuations, some situations may give people a day
329 or two to prepare while others might call for immediate actions. To prepare
330 for evacuations, possible problems should be carefully considered and coun-
331 termeasures should be developed. For example, contraflow operations can be
332 implemented for hurricane evacuations in coastal areas to move the most traf-
333 fic towards inland safety, and AI methods can help practical implementations
334 by determining when to activate contraflow lane reversals (Burris et al. 2015).
335 While large crowds move in different routes during evacuations, it is necessary
336 to estimate crowd dynamics (Jiang et al. 2017; Wang et al. 2019b; Zheng and
337 Liu 2019), identify the best evacuation paths (Peng et al. 2019), and develop
338 evacuation support systems (Higuchi et al. 2014). The most popular AI meth-
339 ods applied for evacuations (Application Area 8) include SVM, DT, neural
340 networks, and reinforcement learning, as well as optimization algorithms.

341 3.3 AI Applications in Disaster Response

342 Timely disaster responses are a matter of life and death. Decision-makers need
343 to make best efforts to understand the situation and improve the efficiency
344 of response efforts. This naturally requires situation awareness for effective
345 decision-making (Application Areas 9 and 10) and user-friendly disaster in-
346 formation systems for effective coordination (Application Area 12) to ensure
347 disaster relief and address people’s urgent needs and concerns (Application
348 Areas 11 and 13). AI methods can be applied to facilitate relief and response
 14 Wenjuan Sun et al.

349 efforts. In general, supervised and unsupervised models, and deep learning
350 have been extensively applied to Areas 9 and 10, while other AI methods are
351 rarely adopted for the two areas. Most AI methods have been applied to Area
352 11. Mainly supervised models and deep learning algorithms have been applied
353 to Areas 12 and 13.
354 Developing maps of the impact area(s) is essential for situation awareness,
355 supporting efficient disaster response efforts (Ramchurn et al. 2015, 2016).
356 Event maps and damage information that are generated from different AI
357 methods can provide vital information for planning search and rescue oper-
358 ations, staging and deploying resources, and understanding short-term hous-
359 ing needs (Vieweg 2012; Lin 2015; Kim et al. 2018c; Rizk et al. 2019). Huge
360 volumes of disaster-related data are continuously generated from satellites
361 (Eguchi et al. 2008), unmanned aerial vehicles (Aljehani and Inoue 2018),
362 robots (Park et al. 2019), and social media (Cervone et al. 2016), based on
363 which disaster event maps can be generated. For instance, satellite images have
364 been used to generate maps of infrastructure inventory models (Eguchi et al.
365 2008), damaged buildings and bridges (Adams et al. 2002; Hutchinson and
366 Chen 2005; Balz and Liao 2010), and disaster-impacted regions (Casagli et al.
367 2017; Rosser et al. 2017). By rapidly analyzing these data with computer vision
368 methods, “live maps” are generated to represent disaster situations (Lucieer
369 et al. 2014; Middleton et al. 2014; Fohringer et al. 2015; Valkaniotis et al. 2018;
370 Xiao et al. 2018). When analyzing maps and images, classifier algorithms are
371 often used (Vetrivel et al. 2016). By comparing maps and images pre-event and
372 post-event, feature discrepancies can be extracted to assess damage of struc-
373 tures and infrastructures for prioritizing response efforts (van Aardt et al.
374 2011; German et al. 2013; Bevington et al. 2015; Koch et al. 2016; Axel and
375 van Aardt 2017; Cresci et al. 2015; Cervone et al. 2016; Nguyen et al. 2017).
376 Different databases have been established for supporting damage assessment
377 for different structures and hazards, such as xBD for building damage assess-
378 ment (Gupta et al. 2019), and HOWAS21 (Kellermann et al. 2020) and FIMA
379 NFIP Redacted Claims Data Set (FEMA 2019) for flood damage assessment.
380 Crowd-sourced information becomes increasing popular in supporting disas-
381 ter response. Many volunteer efforts focus on speeding up the data analysis
382 process to rapidly generate maps and provide invaluable crowdsourced infor-
383 mation for situation awareness and damage assessment (Barrington et al. 2011;
384 Ghosh et al. 2011; Butler 2013). By harnessing “crowds” of over 1000 experts
385 from 82 countries, for example, the Humanitarian OpenStreetMap Team gen-
386 erated devastation maps of the affected areas in the Philippines shortly after
387 typhoon Haiyan, enabling rapid damage assessment and efficient response ef-
388 forts (Butler 2013).
389 In disaster rescue and relief, utilizing social media and robotics as well
390 as mobile phone data often support timely and effective decision-making. So-
391 cial media platforms are powerful communication tools for individuals and
392 local communities to seek help and for governments and organizations to dis-
393 seminate disaster relief information (Li and Rao 2010; Tatsubori et al. 2012;
394 Takahashi et al. 2015). Social media data embed time and geo-location in-
 Applications of artificial intelligence for disaster management 15

395 formation as well as disaster-related information, serving as good information

396 sources for building disaster information systems (Goodchild and Glennon
397 2010; Srivastava et al. 2012; Laylavi et al. 2017). This ultimately supports
398 decision-making for disaster relief and resource allocations (Castellanos et al.
399 2018) and for building disaster information systems (Aydin and Fellows 2018).
400 To analyze social media data, popular AI methods include classifiers, reinforce-
401 ment learning, deep reinforcement learning, and other sentiment analysis tech-
402 niques. However, there are concerns of using social media data as information
403 sources due to issues of credibility, reliability, and difficulties in verifying infor-
404 mation and processing big data into actionable knowledge (Acar and Muraki
405 2011; MacEachren et al. 2011; Tapia et al. 2011).
406 In the aftermath of a disaster, the harsh environment hinders human ef-
407 forts of disaster rescue. Disaster robots allows responders and stakeholders to
408 sense and act at a distance from the impacted areas (Murphy 2014). Robots
409 can serve as remote sensing platforms for mapping and interacting with the
410 destroyed environment (Adams et al. 2014; Kochersberger et al. 2014; Stefanov
411 and Evans 2014), fight fires in dangerous conditions (Schneider and Wilder-
412 muth 2017; Ando et al. 2018), search and rescue (Murphy and Stover 2007;
413 Murphy et al. 2009; Steimle et al. 2009; Zhang et al. 2014; Bakhshipour et al.
414 2017; Hu et al. 2019), and inspect damage (Devault 2000; Murphy et al. 2011;
415 Torok et al. 2014; Ellenberg et al. 2015; Lattanzi and Miller 2015, 2017). Ma-
416 chine learning has been widely used for robotics to acquire new skills and adapt
417 to the surrounding environment (Lenz 2016). For example, deep learning has
418 been applied to visual detection (Socher et al. 2008; Giusti et al. 2015), han-
419 dling multiple input data (Ngiam et al. 2011; Noda et al. 2014), and robotic
420 manipulation (Saxena et al. 2008; Gemici and Savena 2014; Lenz 2016). In ad-
421 dition, optimization algorithms are often used for dynamic path planing and
422 multi-robot communication and coordination (Liu et al. 2013; Takeda et al.
423 2014).
424 One of the first things people commonly do during a disaster is to contact
425 emergency services (and loved ones). Therefore, telecommunications volume
426 sharply increases, usually following the jump-delay pattern (Bagrow et al.
427 2011). In disaster response, disaster management agencies need to rapidly
428 classify information from such calls and share urgent needs of the public to
429 relevant agencies and utility companies. Machine listening can help to au-
430 tomatically recognize voices to identify key words with a high priority and
431 rapidly process voice data from different regions (Ramchurn et al. 2016). With
432 natural language processing algorithms, sentiment mining can help disaster
433 managers perform crisis management and enable efficient disaster relief with
434 better awareness of the situation, such as where to send first responders and
435 distribute resources. Based on the location information of the nearby commu-
436 nication network mast, mobile phone data have also been used to estimate
437 population movements and track population displacement in the immediate
438 aftermath of disasters (Gonzalez et al. 2009; Tatem et al. 2009; Bengtsson
439 et al. 2011). Oftentimes, disasters may completely destroy the base stations of
440 the mobile communication network, and so alternative base stations should be
 16 Wenjuan Sun et al.

441 rapidly established and allocated to support emergency communication, with

442 different countermeasures proposed (Suriya and Sumithra 2019; Wang et al.
443 2019d; Samir et al. 2019).
444 Information sharing and coordination is often the bottleneck in multi-
445 agency response due to the unpredictable and dynamic nature of the dis-
446 aster environment (Chen et al. 2008a,b). As the disaster unfolds, the informa-
447 tion of the disaster event and its impact, victims, and resources may become
448 outdated with large uncertainty and unpredictability by the time of sharing,
449 making life-and-death decision-making very challenging (Holguı́n-Veras et al.
450 2012). Disaster information systems with shared access across agencies and
451 organizations can help address these issues, such as collaborative geographic
452 information systems (Sun and Li 2016; Abdalla and Esmall 2018; Li et al.
453 2019c), shared information management platforms (Bunker et al. 2015; Ra-
454 souli 2018) and decision tools (Moskowitz et al. 2011). With the shared data,
455 collaborative data analytics can be implemented to learn about the disaster
456 situation and identify relief needs (Tucker et al. 2017). Disaster information
457 systems with automatic data-sharing capacity can help decision-makers from
458 different organizations coordinate response efforts in a timely manner. Such
459 ideas have been implemented in the forms of various prototypes (Bartoli et al.
460 2013; Lin and Liaw 2015; Foresti et al. 2015; Kim et al. 2018a; Hochgraf et al.
461 2018). There are multiple applications for disaster information systems by us-
462 ing supervised models and deep learning to extract information from social
463 media data (Neppalli et al. 2018), mobile phone data (Sun and Tan 2019),
464 remote sensing data and aerial images (Morito et al. 2016; Tian and Chen
465 2017b). Example disaster information systems include MADIS (Yang et al.
466 2012), Sahana (Careem et al. 2006), SPIDER (S̆ubik et al. 2010), CrowdHelp
467 (Besaleva and Weaver 2013), and DMCsim (Hashemipour et al. 2017).
468 A disaster causes not only physical damage to structures and infrastructure
469 but also mental damage to people. Different types of feelings will make human
470 focus their attention on very different information and lead to completely dif-
471 ferent decisions and actions (Watson and Clark 1994; Greifeneder et al. 2011).
472 Understanding feelings and psychological needs of victims would be helpful for
473 effective disaster relief (Lin et al. 2017b; Li et al. 2019a). AI methods can help
474 in this regard by analyzing social media data to track feelings and reactions
475 of the public. Social media data embed emotional text and images, time and
476 geo-location information, which as useful to identify the spatial and temporal
477 evolution of public behaviors and population mobility, as well as psychological
478 and healthcare needs (Bengtsson et al. 2011; Caragea et al. 2014; Ukkusuri
479 et al. 2014; Wilson et al. 2016; Kuang and Davison 2017). Previous studies
480 show that there are human activity abnormalities in the physical proximity of
481 the disaster event with obvious spatial and temporal disparities (Chae et al.
482 2014; Shelton et al. 2014; Kryvasheyeu et al. 2016; Neppalli et al. 2017; Liu
483 et al. 2019b; Zou et al. 2019). There are many research efforts working on this
484 area (Area 13), such as developing metrics with sentiment analyses to quantify
485 people’s reaction/emotion in response to response efforts (Neppalli et al. 2017;
486 Bhavaraju et al. 2019; Singh et al. 2019; Chen et al. 2020).
Applications of artificial intelligence for disaster management 17

Table 3 Example AI applications for disaster response

AI Method 9. Event mapping 10. Damage assessment 11. Disaster rescue and 12 Disaster information 13. Understanding
relief, resource system and people’s concern,
allocation collaboration emotion and reaction
A. Linear regression NA NA Bagloee et al. (2019) NA NA
B. Non-linear NA NA Liang et al. (2001); NA NA
regression Luo et al. (2013);
Robinson et al. (2014)
C. Logistic regression Yang and Cervone NA Zhang et al. (2010); Jia NA Gopnarayan and
(2019) and Zhang (2012); Deshpande (2019); Yu
Hung et al. (2016); et al. (2019)
Reynard and Shirgaokar
(2019)
D. Support vector Moskowitz et al. Tan et al. (2010); Kiatpanont et al. Maharjan et al. (2018) Yu et al. (2019);
machine (2011); Ilyas (2014); Ashktorab et al. (2014); (2016); Basu et al. Gopnarayan and
Cresci et al. (2015); Izadi et al. (2017); (2019a); Chaudhuri Deshpande (2019); Ruz
Ireland et al. (2015); Pogrebnykov and and Bose (2020) et al. (2020)
Jiang and Friedland Maldonado (2017);
(2016); Yang and Naito et al. (2018);
Cervone (2019) Zhang et al. (2018a);
Seydi and Rastiveis
(2019)
E. Naı̈ve Bayes Ilyas (2014); Li et al. Imran et al. (2013); Kiatpanont et al. Neppalli et al. (2018) Verma et al. (2011)
(2018a) Mangalathu et al. (2016); Yoon et al.
(2019) (2016); Basu et al.
(2019a)
F. Decision tree Bahrepour et al. Mangalathu et al. Kiatpanont et al. Barrientos and Sainz NA
(2010); Yang and (2019) (2016); Berawi et al. (2012)
Cervone (2019) (2019)
G. Random forest Feng et al. (2019); Conner et al. (2016); Acuna et al. (2017) NA Ruz et al. (2020)
Yang and Cervone Mangalathu et al.
(2019) (2019); Kellermann
et al. (2020)
H. K-nearest neighbor Kim et al. (2016b); Mangalathu et al. Kiatpanont et al. NA Gopnarayan and
Zhao et al. (2019) (2019) (2016); Liu et al. Deshpande (2019)
(2019a)
I. Logistic model tree NA NA Ahmad et al. (2017) NA NA
J. Neural networks Yu et al. (2005); Bandara et al. (2014); Bayerlein et al. (2018); Datt et al. (2015); NA
Kovordányi and Roy Conner et al. (2016); Chaudhuri and Bose Tian and Chen (2017b)
(2009); Yang and Rudner et al. (2019) (2020)
Cervone (2019)
K. Hierarchical Middleton et al. Zhou et al. (2017b) Guha et al. (1998); Zheng et al. (2011, Lodree and Davis
clustering (2014) Kondaveti and Ganz 2013a); Li et al. (2016)
(2009) (2016b)
L. K-means clustering Ganesan et al. (2016) Atasever (2017); Hou ZIDI et al. (2019) NA NA
et al. (2017)
M. Fuzzy clustering Wang et al. (2012); Tan et al. (2010); Yu Sheu (2007, 2010); NA NA
Ganesan et al. (2016) and Zhu (2014); Zeng Ruan et al. (2016)
et al. (2018)
N. Principal NA Hutchinson and Chen Basu et al. (2019b) NA NA
component analysis (2005); Bandara et al.
(2014); Zhou et al.
(2018b)
O. Hidden Markov Salmane et al. (2015) NA Suganya and Jayashree Qiu et al. (2014) NA
models (2018)
P. Convolutional Kim et al. (2016c); Alam et al. (2017); Basu et al. (2019a); Neppalli et al. (2018); Yu et al. (2019); Li
neural networks Liu and Wu (2016); Kamilaris and Boldú Hartawan et al. (2019); Kumar et al. (2020) et al. (2016a)
Bejiga et al. (2017); (2017); Nguyen et al. Robertson et al.
Kamilaris and Boldú (2017); Tian et al. (2019); Chaudhuri and
(2017); Lee et al. (2018); Vetrivel et al. Bose (2020)
(2017); Huang et al. (2018); Xu et al.
(2019c,b); Ahmad (2019a); Zhang et al.
et al. (2019) (2019a); Pogrebnykov
and Maldonado (2017);
Seydi and Rastiveis
(2019)
Q. Recurrent neural Kundu et al. (2018); Nguyen et al. (2019b); NA Neppalli et al. (2018); Hernandez-Suarez et al.
networks Mao et al. (2019); Moustapha and Selmic Kumar et al. (2020) (2019)
Rahnemoonfar et al. (2007); Verma et al.
(2018) (2020); Biswas et al.
(2019); Pogrebnykov
and Maldonado (2017)
R. Deep neural Khan et al. (2017); Bai et al. (2018) NA Morito et al. (2016); NA
network Bai et al. (2018) Neppalli et al. (2018)
S. Multi-layer NA Seydi and Rastiveis Robertson et al. (2019) NA NA
perception (2019)
T. Recursive neural NA NA NA NA Dong et al. (2014)
network
U. Q-learning NA Zhao et al. (2017) Su et al. (2011); Qiao and Luo (2012); NA
Castellanos et al. Aydin and Fellows
(2018); Liu et al. (2018)
(2019a); Hou et al.
(2019)
V. Policy gradient NA Mao et al. (2016); Rodriguez-Ramos et al. NA NA
Wang et al. (2019e) (2019); Silver et al.
(2014)
W. Deep Q-networks Baldazo et al. (2019); Maciel-Pearson et al. Wang et al. (2020a); Huang et al. (2017); NA
Maciel-Pearson et al. (2019) Yang and Liu (2018); Sun and Tan (2019);
(2019) Guo et al. (2019) Liu et al. (2018)
X. Genetic algorithm NA Izadi et al. (2017); Tian Pessin et al. (2009); NA NA
et al. (2018) Zhao et al. (2009);
Wang (2018); Liu et al.
(2019a); ZIDI et al.
(2019)
Y. Particle swarm NA Xu et al. (2019b) Pugh and Martinoli NA NA
optimization (2007); Sánchez-Garcı́a
(2019); ZIDI et al.
(2019)
Z. Simulated annealing NA NA Fiedrich et al. (2000); NA NA
Yadollahnejad et al.
(2017); ZIDI et al.
(2019)
18 Wenjuan Sun et al.

Table 4 Example AI applications for disaster recovery

AI Method 14. Assess impact 15. Develop recovery 16. Track recovery 17. Evaluate loss and
plan repair cost
A. Linear regression McCaslin et al. NA Zobel (2014); Qiang Barthel and Neumayer
(2005); Zhang and et al. (2020) (2012); Yu et al.
Peacock (2009); (2014); Kim et al.
Rosellini et al. (2018) (2016a); Kousky and
MichelKerjan (2015)
B. Non-linear Haraoka et al. (2012); NA Zobel (2014); Zhang Smith and Katz
regression Mitsova et al. (2018); (2016); Wang et al. (2013); Kim et al.
Rosellini et al. (2018); (2018b); Jamali et al. (2015, 2018b)
Cheng and Zhang (2019); Yabe and
(2020) Ukkusuri (2019); Qiang
et al. (2020)
C. Logistic regression Tunusluoglu et al. NA Gopnarayan and NA
(2007); Nabian and Deshpande (2019)
Meidani (2018a);
Mitsova et al. (2019)
D. Support vector Gong et al. (2013); Oh et al. (2006) Yabe and Ukkusuri NA
machine Nabian and Meidani (2019); Pogrebnykov
(2018a); Moya et al. and Maldonado (2017);
(2018); Rosellini et al. Gopnarayan and
(2018); Sheykhmousa Deshpande (2019)
et al. (2019); Zhang
and Burton (2019)
E. Naı̈ve Bayes NA NA Shibuya and Tanaka NA
(2019)
F. Decision tree Merz et al. (2013); NA NA Stojadinovic et al.
Rosellini et al. (2018) (2017)
G. Random forest Rosellini et al. (2018); NA NA NA
Zhang et al. (2018b)
H. K-nearest neighbors Khaloo et al. (2017); NA Gopnarayan and NA
Moya et al. (2018); Deshpande (2019)
Nabian and Meidani
(2018a)
I. Logistic model tree NA NA NA NA
J. Neural networks Mehrjoo et al. (2008); Asgary and Naini (2011) NA Chen and Huang
Khoshnoudian et al. (2006);
(2017); Padil et al. Aghamohammadi et al.
(2017) (2013)
K. Hierarchical NA NA NA NA
clustering
L. K-means clustering NA NA NA NA
M. Fuzzy clustering Yu et al. (2016) NA NA NA
N. Principal Yu et al. (2016); Cha NA NA NA
component analysis and Buyukozturk
(2015); Khoshnoudian
et al. (2017);
Yamaguchi and
Shirota (2019)
O. Hidden Markov NA NA NA NA
models
P. Convolutional Cha et al. (2017); NA Yang et al. (2019c); NA
neural networks Liang (2018); Pogrebnykov and
Ghaffarian et al. Maldonado (2017)
(2019)
Q. Recurrent neural NA NA Pogrebnykov and NA
networks Maldonado (2017)
R. Deep neural Fallahian et al. (2018) NA NA NA
network
S. Multi-layer NA NA Lin et al. (2008) NA
perception
T. Recursive neural NA NA NA NA
network
U. Q-learning NA Memarzadeh and Pozzi NA NA
(2019)
V. Policy gradient NA NA NA NA
W. Deep Q-networks NA Joo et al. (2019); Ning NA NA
et al. (2019); Geng
(2019)
X. Genetic algorithm Alfaiate et al. (2007); Xu et al. (2007); NA NA
Meruane and Heylen Bocchini and Frangopol
(2011); Gomes et al. (2012a,b); Tapia and
(2019) Padgett (2015);
Karamlou and Bocchini
(2016); Eid and
El-adaway (2017a,b); Li
and Teo (2018)
Y. Particle swarm Huang et al. (2019a) NA NA NA
optimization
Z. Simulated annealing Strauss et al. (2009) Hackl et al. (2018) NA NA
 Applications of artificial intelligence for disaster management 19

487 3.4 AI Applications in Disaster Recovery

488 Disaster recovery is a multifaceted process, involving governments and pub-

489 lic authorities, as well as private organizations. This requires comprehensive
490 decision-making to quickly understand the complexity of the situation, identify
491 operational needs and recovery plans, and perform rehabilitation and recon-
492 struction activities. As disaster recovery usually takes a long time, including
493 precise damage assessment, budgeting, planning, permitting, design and con-
494 struction, AI can be an important module for supporting disaster recovery
495 management in less time. AI methods have been applied to disaster recovery
496 management, by assessing the disaster induced impact in detail (Application
497 Area 14), developing recovery plans (Application Area 15), tracking the recov-
498 ery process (Application Area 16), and estimating loss and repair cost (Appli-
499 cation Area 17). The increasing number of publications in recent years, shown
500 in Table 4, indicates increasing attention to applying AI for disaster recovery
501 management. Among them, more attention has been paid to Application Area
502 14 than others (Application Areas 15, 16 and 17).
503 Quick and accurate assessment of the disaster-induced impact is critical for
504 rapid recovery. In addition to physical damage, a disaster causes psychological
505 distress and economic disturbance. When assessing physical damage, visual in-
506 spection is a primary method adopted in current practice for buildings (Pham
507 et al. 2014; Choi et al. 2018; Lenjani et al. 2019), bridges (Yeum and Dyke
508 2015), tunnels (Victores et al. 2011), storage tanks (Schempf et al. 1995), etc.
509 However, the visual inspection method is often tedious and labor intensive.
510 AI methods can help eliminate such human efforts based on aerial images,
511 social media imagery data, and sensor measurement data (Khaloo et al. 2017;
512 Khoshnoudian et al. 2017). When assessing the disaster-induced impact on
513 human, sentiment analyses of social media data can track human activity pat-
514 tern throughout the recovery (Caragea et al. 2014; Hasan and Ukkusuri 2014;
515 Shelton et al. 2014; Resch et al. 2018; Liu et al. 2019b). When investigat-
516 ing psychological distress following a disaster, the use of surveys is a primary
517 method adopted in current practice. Both supervised and unsupervised mod-
518 els, particularly regression methods, dimension reduction methods, and neural
519 networks, are often adopted to analyze survey results to identify risk factors
520 and assess the effectiveness of preventive interventions (Gao et al. 2006; Kim
521 et al. 2008; Huang et al. 2010; Gong et al. 2013; Rosellini et al. 2018). In
522 addition, AI methods have been applied to estimate the economic impacts of
523 a hazard, in which supervised models are often used to establish quantitative
524 relations between critical factors and the economy and identify possible stim-
525 ulus for economic growth (Zhang and Peacock 2009; Yamaguchi and Shirota
526 2019; Cheng and Zhang 2020; Qiang et al. 2020).
527 After precisely assessing the disaster induced impact, establishing post-
528 event recovery plans is essential for effectively conducting recovery and re-
529 newal activities. While pre-event planning allows participation members to
530 spend significant time and resources for fostering cooperative plans, post-event
531 planning is often carried out in a relatively hostile environment with less time
 20 Wenjuan Sun et al.

532 and resources at hand. In current research, optimization techniques are often
533 adopted to identify efficient plans of restoration, or to estimate human deci-
534 sions of recovery planning (Sun et al. 2021), including genetic algorithms (Xu
535 et al. 2007; Orabi et al. 2010; Bocchini and Frangopol 2012b; Karamlou and
536 Bocchini 2016), and simulated annealing (Hackl et al. 2018), and other meth-
537 ods (Sarkale et al. 2018; Zhong et al. 2018). Additionally, there are few studies
538 applying reinforcement learning and deep reinforcement learning to planning
539 post-event recovery strategies (Joo et al. 2019; Ning et al. 2019).
540 During the recovery process, practitioners need metrics and tools to mea-
541 sure and monitor how well a community recovers from a disaster over time
542 as a means of building community resilience (Curtis et al. 2007). Supervised
543 models and deep learning algorithms are often used in this aspect by analyz-
544 ing data from various sources. As social media data are attached with geotags
545 or hashtags, using sentiment analysis methods and image classification tech-
546 niques to analyze social media data can be very helpful for disaster recovery
547 tracking (Eckle et al. 2017; Pogrebnykov and Maldonado 2017; Jamali et al.
548 2019; Malawani et al. 2020; Mihunov et al. 2020). By comparing nighttime
549 light data at different time, established regression relations between economic
550 indicators and spatial variations in light intensity can provide valuable insights
551 about how the regional economy recovers in a quantitative manner (Wang et al.
552 2018b; Qiang et al. 2020). Using Google Street View to remotely track disas-
553 ter recovery has also become increasingly popular (Curtis et al. 2010; Mabon
554 2016).
555 In the aftermath of a disaster, governments need to provide timely assis-
556 tance to reconstruct homes and rebuild lives; there are urgent demands for a
557 rapid assessment of loss estimate and repair cost (Eguchi et al. 1998; Ladds
558 et al. 2017; Deryugina 2017). AI methods can help estimate disaster losses and
559 repair costs. In particular, supervised models, such as regression and neural
560 network, have been used to rapidly process imagery for detecting structural
561 damage, identifying repair needs, and estimating repair cost; they have also
562 been used to analyze historical dispersion data of disaster recovery funds for
563 budget allocations, and process insurance claims in less time (Chen and Huang
564 2006; Barthel and Neumayer 2012; Zagorecki et al. 2013; Stojadinovic et al.
565 2017). The existence of only a small number of publications in this field indi-
566 cates that AI applications to Area 17 is still in its infancy. In current practice,
567 the disaster loss and repair cost are usually estimated based on real data from
568 different sources, such as insurance claims, post-disaster assessment, and as-
569 sistance grants and personal loans to victims (Eguchi et al. 1998; Kim et al.
570 2015). The availability of big data and the rapid development of data analyt-
571 ics offer an unprecedented opportunity to promote AI applications in rapid
572 estimation of disaster loss and repair cost in the near future. However, the
573 lack of standardized methods for collecting and recording data may lead to
574 very different estimates of economic impacts (Ladds et al. 2017). Therefore,
575 establishing policies and standards for data collection is an urgent need.
576 After a disaster, disaster related rumors and fraud may appear, requiring
577 the awareness and alertness of both disaster victims and governments. Data
 Applications of artificial intelligence for disaster management 21

578 mining can help to identify potential fraud (Bagde and Chaudhari 2016; Dutta
579 et al. 2017) and rumors (Mendoza et al. 2010; Liu et al. 2015; Wu et al. 2015;
580 Zubiaga et al. 2016, 2018), as well as track trends of information flow (Hong
581 et al. 2011; Badmus 2020). For example, insurance companies and law enforce-
582 ment agencies can use machine learning to quickly examine the truthfulness
583 of a claim for a flooded house by making a before-and-after comparison of
584 high-resolution satellite images (Gilmour 2019).

585 4 Practical AI-based Decision Support Tools

586 To ultimately facilitate informed disaster management in practice, many AI-

587 based decision support tools have been developed by research institutes and
588 industrial companies in the past few decades. By searching on websites of
589 Google Scholar and Web of Science with keywords of “disaster management”,
590 “decision support tool”, and “artificial intelliegence”, we have found related
591 AI-based tools for decision-making in disaster management. Table 5 presents
592 example tools that apply various AI techniques in disaster management. These
593 tools make use of various data as input to extract useful information, including
594 social media data, mobile phone data, sensor measurements, on-site reports
595 from first responders, and crowdsourced information from volunteers. These
596 tools cover different infrastructures and different types of hazards, contributing
597 to the advancement of AI applications to fostering informed disaster manage-
598 ment at different phases. A general trend is that there are more tools applicable
599 for the disaster response phase than other phases. Most tools use social me-
600 dia data as input; a small portion of tools use sensor measurements, remote
601 sensing data, or mobile phone data as input.
602 Some tools focus on predicting possible consequences under a hazard sce-
603 nario for developing management plans of retrofit and evacuation in the dis-
604 aster mitigation and preparedness phases. For instance, Optima predictT M
605 software simulates and predicts emergent medical service demand and ambu-
606 lance availability changes in the wake of a disaster, helping dispatchers and
607 operations personnel find possible optimal ways of preparing for unexpected
608 emergencies (Mason 2013). Other tools provide comprehensive platforms for
609 efficient communications with text, audio, and location services for professional
610 response teams in the disaster response phase, as saving life is typically the
611 most critical issue in the first few days after a disaster and requires commu-
612 nication and situational awareness (Yin et al. 2012b). For example, Blueline
613 Grid analyzes real-time mobile phone data for efficient disaster responses. One
614 Concern predicts possible infrastructure damages and consequences based on
615 infrastructure data and historical disaster data. Artificial Intelligence for Dis-
616 aster Response (AIDR) automatically classifies crisis-related tweets along with
617 crowdsourced information of aerial images to identify victims’ needs and infras-
618 tructure damage for efficient disaster response management (Imran et al. 2014;
619 Ofli et al. 2016). SensePlace3 is a geo-visual interface that can visualize time,
620 location, and relationships of events, by applying data mining tools available
 22 Wenjuan Sun et al.

621 in Solr to process real-time Twitter data (Tomaszewski et al. 2011; Pezanowski
622 et al. 2018). DeepMob simulates human behavior and mobility during natural
623 disasters by learning from millions of users’ GPS records with deep belief net-
624 works (Song et al. 2017). GeoQ is an open-source tool for assessing damage by
625 crowdsourcing geo-tagged photos of the disaster-affected areas, developed in
626 coordination with the National Geospatial-Intelligence Agency, the Presiden-
627 tial Innovation Fellow Program, the Federal Emergency Management Agency
628 (FEMA), and other analysts.
629 In the meantime, there are some challenging issues of using these AI-based
630 decision support tools in practice. First, these tools typically require large
631 amounts of data as input, and data-related issues are a practical challenge.
632 Input data might be available in different types and formats for different com-
633 munities, or available for some communities but not available for others due to
634 various reasons, such as legal ramifications and commercial competitiveness.
635 For example, big cities and urban areas usually have documented data detailed
636 enough and sufficient in size to make AI predictions accurate, which may not
637 be the case for small cities and rural areas. Even if all input data are avail-
638 able, some of it may be inaccurate, and there may be data ownership issues
639 involved when using some of these tools. Therefore, policies and regulations
640 need to be established for appropriate data collection, cleaning, protection,
641 and management. Second, communities are exposed to different types of haz-
642 ards and have different socioeconomic backgrounds. The AI-based decision
643 support tools that are developed based on data from one community might
644 not be suitable for another community. This naturally poses a challenge to
645 the application generalization of AI-based decision support tools for a diverse
646 set of communities. Third, some tools may require a high level of competence
647 in deployment, making them less user friendly for practitioners. Many tools
648 require advanced software and high performance computers to conduct big
649 data analytics, which may not be available for many local governments and
650 emergency agencies in economically disadvantaged regions.

651 5 Discussion

652 As shown in Tables 1 ∼ 4, all AI methods have been applied to disaster man-
653 agement. However, there are many untouched application areas by some AI
654 methods. For instance, very few AI methods have been used for disaster train-
655 ing systems (Application Area 7); that is probably because there is very little
656 training data of human responses in disasters available to build appropriate
657 AI models for such purposes. Deep neural networks (method R) and recursive
658 neural networks (method T ) are rarely applied for disaster preparedness and
659 disaster recovery (Application Areas 5 ∼ 8 and 14 ∼ 17). Policy gradient-based
660 algorithms have not been applied in disaster mitigation and disaster recovery
661 (Application Areas 1 ∼ 4 and 14 ∼ 17). The absence of AI applications to
662 untouched areas may attract future research attention for exploration.
Applications of artificial intelligence for disaster management 23

Table 5 AI-based decision support tools for disaster management

Example tool Owner Input data Hazard Applicable phase Website / Reference
Mobile phone data,
Optima https://www.r1rcm.
Intermedix clinical data, and General Mitigation
PredictT M com/optima
others
Public and private
Seismic, Mitigation, and https://www.
One Concern One Concern, Inc. infrastructure
flood response oneconcern.com
data-sets
Data of hazard
The Geospiza Mitigation, and https://geospiza.us/
Geospiza Inc. modeling, community, General
Solution response solution
and live event
Arizona State Preparedness, and http://tweettracker.
TweetTracker Tweet General
University response fulton.asu.edu/
National Research
EARS Twitter Earthquake Preparedness Avvenuti et al. (2014)
Council, Italy
University of McCreadie et al.
EAIMS Twitter General Preparedness
Glasgow (2016)
Sandia National Human decision input Djordjevich et al.
Ground Truth General Preparedness
Laboratories via video games (2008)
Preparedness, and
Argus Rutgers University Smartphone data General Sadhu et al. (2017)
response
https:
Crisis Mappers Preparedness, and
CrisisMappers Social media data General //crisismapping.
Net response
ning.com/
Preparedness, and https://www.
Dataminr Dataminr Social media data General
response dataminr.com/
Disaster
Management Infrastructure data,
George
Coordination GIS data, and Preparedness, and Hashemipour et al.
Washington General
simulation organization response (2017)
University
(DMCsim) capabilities
system
Artificial
Intelligence
Qatar Computing
for Digital Tweets General Response http://aidr.qcri.org
Research Institute
Response
(AIDR)
https://www.
Blueline Grid WorldAware, Inc Mobile phone calls General Response
bluelinegrid.com
https://www.
Blueworx Blueworx Emergency calls General Response
blueworx.com
Stanford
CRED Seismogram data Earthquake Response Mousavi et al. (2019)
University
Disaster data, human
Multi-
mobility data,
government-
DeepMob earthquake records, Earthquake Response Song et al. (2017)
industry
transportation
collaborations
network data
Information
Information
ESA Engineering General Response Yin et al. (2012a)
management system
Laboratory
University of
Southampton,
University of Social media data and Ramchurn et al.
HAC-ER General Response
Nottingham, and first responder reports (2015, 2016)
University of
Oxford
Pennsylvania Pezanowski et al.
SensePlace3 Tweets General Response
State University (2018)
Sahana Information
Sahana General Response Careem et al. (2006)
Foundation management system
Mitigation,
Disaster Images, data of
Disaster preparedness, https://www.
Intelligence hazard, infrastructure, General
Intelligence response, and disaster-ai.com
product and community
recovery
Mitigation,
Remote sensing data
Disaster City Texas A&M preparedness,
and crowd-sourced General Fan et al. (2019)
Digital Twin University response, and
data
recovery
Federal
Disaster Emergency Photos and Response, and https://www.fema.
General
Reporter Management descriptive text recovery gov/disaster-reporter
Agency
Florida Preparedness,
Zheng et al. (2013a);
FIU-Miner International Geospatial data General response, and
Li et al. (2017a,b)
University recovery
National
Geospatial- Response, and https://github.com/
GeoQ Geo-tagged photos General
Intelligence recovery ngageoint
Agency
Tweet
United States Response, and https://github.com/
Earthquake Tweets Earthquake
Geological Survey recovery usgs/earthquake-ted
Dispatch
Flood, fire,
Tractable Tractable Images Recovery https://tractable.ai
hurricane
 24 Wenjuan Sun et al.

663 Many challenges of practical AI applications to disaster management are

664 due to data-related issues, such accessibility, completeness, security, privacy,
665 and ethical issues (Boyd and Crawford 2012; Crawford and Finn 2015). Mak-
666 ing accurate predictions with AI techniques typically requires a large amount
667 of good data for building the model. Such data is not always available. For
668 example, some infrastructure data cannot be easily accessible due to reasons
669 of national security and commercial competitiveness. Data trustworthiness
670 is another issue. For instance, raw data from social networks often contain
671 various inaccuracies and biases, requiring advanced information filtering and
672 verification. One step further, collecting and analyzing personal data poses
673 significant issues related to fairness, responsibility, and human rights. Even if
674 the required data are available, data incompleteness is a common problem in
675 disaster-related data analyses due to the dynamically changing environment
676 of a disaster. To deal with the aforementioned issues, there have been various
677 platforms and databases built to collect and share disaster-related data in a rel-
678 atively standardized form. Some examples include ShakeMap and ShakeCast
679 (USGS 2016b,a), GeoPlatform (GeoPlatform 2016), I-WASTE (EPA 2016),
680 Lantern Live (DOE 2014), and Disaster Response Program (ESRI 2016), De-
681 signSafe (NHERI 2019), xBD (Gupta et al. 2019), etc.
682 There are three computation-related challenging issues. First, there may
683 not be enough human labelled training data in time considering the increas-
684 ing amount of data and the limited amount of manpower in the wake of a
685 disaster (Pouyanfar et al. 2018). In this regard, applying and improving un-
686 supervised learning approaches may be the way out for handling real-world
687 data without manual human labels (Ranzato et al. 2013). Second, the compu-
688 tational complexity sharply increases with the size, variety, and update rate
689 of data, which challenges the capacity of processing, managing, and learning
690 data within a reasonable response time in the disaster scenario. Efficiently
691 managing, storing, and processing big data is essential for disaster manage-
692 ment, particularly disaster response. Using cloud platforms to efficiently query
693 and store big data is helpful to address this challenge. Developing more effi-
694 cient AI methods would naturally be helpful. There have been efforts made to
695 address this challenge, including reservoir computing (Tanaka et al. 2019) and
696 using GPUs and AI accelerators (Wang et al. 2019f). Using crowd-sourcing
697 with real-time AI analyses can help to complete the necessary computation
698 within the time limit and eliminate the amount of necessary but tedious work
699 that traditionally needs effort on-site (Bevington et al. 2015). Third, building
700 user-friendly tools for disaster management is essential for practitioners. This
701 means building AI-based tools with interfaces that require minimal technical
702 expertise for practical use.
703 Analysis results from AI models should be explainable and repeatable for
704 supporting practical disaster management. To address this issue, there have
705 been research efforts made to improve the interpretability and explainability of
706 AI models, such as explainable artificial intelligence (Arrieta et al. 2020; Gun-
707 ning et al. 2019). On the other hand, as AI solutions are developed for disaster
708 management, we recognize that there are often challenges in reproducibility of
 Applications of artificial intelligence for disaster management 25

709 new results. For disaster related data, the non-reproducibility issue is a par-
710 ticular challenge, because disasters happen irregularly with various impacts
711 in different regions (Wang et al. 2016). Replication of experimental results is
712 essential for trustworthy advancement in science generally and for AI mod-
713 els specifically. To address this issue, there have been research efforts such as
714 IBM’s AI OpenScale and OpenML (Vanschoren et al. 2014; Rossi 2019; Yang
715 et al. 2019a). These efforts work toward making AI transparent and trust-
716 worthy by capturing the processes, data, and parameters for experiments to
717 become repeatable.

718 6 Concluding Remarks

719 This study focuses on AI applications in assisting in efficient disaster man-

720 agement during four disaster management phases: mitigation, preparedness,
721 response, and recovery. In particular, this study reviews applications of a total
722 of 26 AI methods in 17 Application Areas in disaster management in all four
723 phases. Both research and practice show that analysis results from AI models
724 are very useful for supporting disaster management. In the current stage, the
725 general trend is that most applications focus on disaster response, followed by
726 disaster mitigation.
727 AI is better than humans in terms of data analysis speed and thus the
728 volume of analyzable data. It can make acceptable forecasts when the scope
729 is within the range of the training data, but predictions when the scope is
730 beyond the range may be unacceptable. This is especially true as both the
731 hazard and the society are constantly evolving, which might fundamentally
732 change the utility of attributes used to train the original model. Even if AI
733 algorithms can make reasonably good predictions with the available data, a
734 further concern is whether we should completely rely on the predictions and
735 suggestions from AI algorithms to deploy resources and develop disaster plans.
736 This question has no simple answer.
737 For practical AI applications in disaster management, there are a number
738 of challenging issues related to data and computation, as well as inseparability
739 and replicability of analysis results. This study also identifies many untouched
740 application areas of different AI methods. How to develop more powerful and
741 cost-effective AI-based tools to support decision-making in practical disaster
742 management with improved analysis accuracy and speed is an urgent problem
743 for the research community. Despite these challenges and untouched areas,
744 AI methods provide numerous opportunities and easy solutions for various
745 successful applications in disaster management. By discussing the application
746 status of AI methods in disaster management, this study aims to inspire fu-
747 ture research to tackle the identified challenging issues and advance disaster
748 management with AI for improving community disaster resilience.
 26 Wenjuan Sun et al.

749 Acknowledgements This work is part of the Probabilistic Resilience Assessment of In-
750 terdependent Systems (PRAISys) project (www.praisys.org). Support from the National
751 Science Foundation through grant CMMI-1541177 is gratefully acknowledged.

752 References

753 Abdalla R, Esmall M (2018) Artificial intelligence and WebGIS for disaster and emergency
754 management. In: WebGIS for Disaster Management and Emergency Response, Springer,
755 pp 57–62
756 Abdullahi SI, Habaebi MH, Malik NA (2018) Flood disaster warning system on the go. In:
757 Proceedings of the 2018 7th International Conference on Computer and Communication
758 Engineering (ICCCE), IEEE, pp 258–263, DOI: 10.1109/ICCCE.2018.8539253
759 Acar A, Muraki Y (2011) Twitter for crisis communication: Lessons learned from Japan’s
760 tsunami disaster. International Journal of Web Based Communities 7:392–402
761 Acuna V, Kumbhar A, Vattapparamban E, Rajabli F, Guvenc I (2017) Localization of WiFi
762 devices using probe requests captured at unmanned aerial vehicles. In: Proceedings of
763 the 2017 IEEE Wireless Communications and Networking Conference (WCNC), IEEE,
764 DOI: 10.1109/WCNC.2017.7925654
765 Adams BJ, Huyck C, Mansouri B, Eguchi R, Shinozuka M (2002) Post-disaster bridge
766 damage assessment. In: Proceedings of the 15th Pecora Conference: Integrating Remote
767 Sensing at the Global, Regional, and Local Scale, 8p, on CD-ROM
768 Adams SM, Levitan M, Friedland CJ (2014) High resolution imagery collection for post-
769 disaster studies utilizing unmanned aircraft systems (UAS). Photogrammetric Engineer-
770 ing & Remote Sensing 12:1161–1168
771 Adeel A, Gogate M, Farooq S, Ieracitano C, Dashtipour K, Larijani H, Hussain A (2018)
772 A survey on the role of wireless sensor networks and iot in disaster management. In:
773 Geological Disaster Monitoring Based on Sensor Networks, Springer, pp 57–66
774 Afandizadeh S, Jahangiri A, Kalantari N (2013) Determination of the optimal network con-
775 figuration for emergency evacuation by simulated annealing algorithm. Natural Hazards
776 69:1315–1335
777 Agarwal S, Kachroo P, Regentova E (2016) A hybrid model using logistic regression and
778 wavelet transformation to detect traffic incidents. IATSS Research 40:56–63
779 Aghamohammadi H, Mesgari MS, Mansourian A, Molaei D (2013) Seismic human loss
780 estimation for an earthquake disaster using neural network. International Journal of
781 Environmental Science and Technology 10:931–939
782 Ahmad K, Riegler M, Pogorelov K, Conci N, Halvorsen P, De Natale F (2017) JORD:
783 A system for collecting information and monitoring natural disasters by linking social
784 media with satellite imagery. In: Proceedings of the 15th International Workshop on
785 Content-Based Multimedia Indexing, ACM, article No. 12
786 Ahmad K, Pogorelov K, Riegler M, Ostroukhova O, Halvorsen P, Conci N, Dahyot R (2019)
787 Automatic detection of passable roads after floods in remote sensed and social media
788 data. Signal Processing: Image Communication 74:110–118
789 Ahmad R, Samy GN, Ibrahim NK, Bath PA, Ismail Z (2009) Threats identification in
790 healthcare information systems using genetic algorithm and cox regression. In: The
791 Fifth International Conference on Information Assurance and Security, pp 757–760,
792 DOI: 10.1109/IAS.2009.313
793 Akhtar N, Siddique B (2017) On hierarchical visualization of event detection in Twitter. In:
794 Advances in Computer and Computational Sciences, pp 571–579
795 Alam F, Imran M, Ofli F (2017) Image4Act: online social media image processing for disaster
796 response. In: Proceedings of the 2017 IEEE/ACM International Conference on Advances
797 in Social Networks Analysis and Mining 2017 (ASONAM‘17), IEEE, pp 601–604
798 Alam P, Booth D, Lee K, Thordarson T (2000) The use of fuzzy clustering algorithm and
799 self-organizing neural networks for identifying potentially failing banks: an experimental
800 study. Expert Systems with Applications 18:185–199
 Applications of artificial intelligence for disaster management 27

801 Alfaiate J, Aliabadi M, Guagliano M, Susmel L (2007) Identification of damaged bars in

802 three-dimensional redundant truss structures by means of Genetic Algorithms. Key En-
803 gineering Materials 348–349:229–232
804 Ali EM, Ahmed MM, Wulff SS (2019) Detection of critical safety events on freeways in
805 clear and rainy weather using SHRP2 naturalistic driving data: Parametric and non-
806 parametric techniques. Safety Science 119:141–149
807 Aljehani M, Inoue M (2018) Safe map generation after a disaster, assisted by an unmanned
808 aerial vehicle tracking system. Transactions on Electric and Electronic Engineering
809 14:271–282
810 Andersson M, Rydell J, St-Laurent L, Prévost D, Gustafsson F (2012) Crowd analysis with
811 target tracking, K-means clustering and hidden Markov models. In: The 15th Interna-
812 tional Conference on Information Fusion, IEEE, pp 1903–1910
813 Ando H, Ambe Y, Ishii A, Konyo M, Tadakuma K, Maruyama S, Tadokoro S (2018) Aerial
814 hose type robot by water jet for fire fighting. IEEE Robotics and Automation 3:1128 –
815 1135
816 Ansari A, Firuzi E, Etemadsaeed L (2015) Delineation of seismic sources in probabilistic
817 seismichazard analysis using fuzzy cluster analysis and Monte Carlo simulation. Bulletin
818 of the Seismological Society of America 105:2174–2191
819 Arridha R, Sukaridhoto S, Pramadihanto D, Funabiki N (2017) Classification extension
820 based on IoT-big data analytic for smart environment monitoring and analytic in real-
821 time system. International Journal of Space-Based and Situated Computing 7:82–93,
822 DOI: 10.1504/IJSSC.2017.10008038
823 Arrieta AB, Daz-Rodrguez N, Ser JD, Bennetot A, Tabik S, Barbado A, Garcı́a S, Gil-López
824 S, Molina D, Benjamins R, Chatila R, Herrera F (2020) Explainable artificial intelli-
825 gence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible
826 AI. Information Fusion 58:82–1155, DOI: 10.1016/j.inffus.2019.12.012
827 Asgary A, Naini AS (2011) Modelling the adaptation of business continuity planning by
828 business using neural networks. Intelligent Systems in Accounting, Finance and Man-
829 agement 18:89–104
830 Ashktorab Z, Brown C, Nandi M, Culotta A (2014) Tweedr: mining Twitter to inform dis-
831 aster response. In: Proceedings of the 11th Proceedings of the International Conference
832 on Information Systems for Crisis Response and Management, ISCRAM, pp 354–358,
833 http://idl.iscram.org/files/ashktorab/2014/275 Ashktorab etal2014.pdf
834 Asim KM, Martı́nez-Álvarez F, Basit A, Iqbal T (2017) Earthquake magnitude prediction
835 in Hindukush region using machine learning techniques. Natural Hazards 85:471–486
836 Atasever UH (2017) A new unsupervised change detection approach with hybrid clustering
837 for detecting the areal damage after natural disaster. Fresenius Environment Bullet
838 26:3891–3896
839 Avvenuti M, Cresci S, Marchetti A, Meletti C, Tesconi M (2014) EARS (earthquake alert
840 and report system): a real time decision support system for earthquake crisis manage-
841 ment. In: Proceedings of the 20th ACM SIGKDD international conference on Knowledge
842 discovery and data mining, pp 1749–1758,
843 Axel C, van Aardt JAN (2017) Building damage assessment using airborne LiDAR. Journal
844 of Applied Remote Sensing 4:046024
845 Aydin ME, Fellows R (2018) Building collaboration in multi-agent systems using reinforce-
846 ment learning. In: International Conference on Computational Collective Intelligence
847 (ICCCI 2018), Springer, pp 201–212
848 Badmus O (2020) When the storm is over: Sentiments, communities and information flow
849 in the aftermath of Hurricane Dorian. International Journal of Disaster Risk Reduction
850 DOI: 10.1016/j.ijdrr.2020.101645
851 Bagde PR, Chaudhari MS (2016) Analysis of fraud detection mechanism in health insurance
852 using statistical data mining techniques. International Journal of Computer Science and
853 Information Technologies 7:925–927
854 Bagloee SA, Johansson KH, Asadi M (2019) A hybrid machine-learning and optimization
855 method for contraflow design in post-disaster cases and traffic management scenarios.
856 Expert Systems with Applications 124:67–81
 28 Wenjuan Sun et al.

857 Bagrow JP, Wang D, Barabási AL (2011) Collective response of human populations to
858 large-scale emergencies. POLS ONE 6:e17680, DOI: 10.1371/journal.pone.0017680
859 Bahrepour M, Meratnia N, Poel M, Taghikhaki Z, Havinga PJ (2010) Distributed event
860 detection in wireless sensor networks for disaster management. In: 2nd International
861 Conference on Intelligent Networking and Collaborative Systems, IEEE, pp 507–512,
862 DOI: 10.1109/INCOS.2010.24
863 Bai SB, Wang J, Lü GN, Zhou PG, Hou SS, Xu SN (2010) GIS-based logistic regression for
864 landslide susceptibility mapping of the Zhongxian segment in the Three Gorges area,
865 China. Geomorphology 115:23–31
866 Bai Y, Gao C, Singh S, Koch M, Adriano B, Mas E, Koshimura S (2018) A framework of
867 rapid regional tsunami damage recognition from post-event TerraSAR-X imagery using
868 deep neural networks. IEEE Geoscience and Remote Sensing Letters 15:43–47
869 Bakhshipour M, Ghadi MJ, Namdari F (2017) Swarm robotics search & rescue: A novel
870 artificial intelligence-inspired optimization approach. Applied Soft Computing 57:708–
871 726
872 Baldazo D, Parras J, Zazo S (2019) Decentralized multi-agent deep reinforcement learning
873 in swarms of drones for flood monitoring. In: 2019 27th European Signal Processing
874 Conference (EUSIPCO), IEEE, DOI: 10.23919/EUSIPCO.2019.8903067
875 Balz T, Liao M (2010) Building-damage detection using post-seismic high-resolution SAR
876 satellite data. International Journal of Remote Sensing 31:3369–3391
877 Bandara RP, Chan TH, Thambiratnam DP (2014) Structural damage detection method
878 using frequency response functions. Structural Health Monitoring 13:418–429
879 Bande S, Shete VV (2017) Smart flood disaster prediction system using IoT & neural net-
880 works. In: 2017 International Conference On Smart Technologies For Smart Nation
881 (SmartTechCon), IEEE, DOI: 10.1109/SmartTechCon.2017.8358367
882 Barabadi A, Ayele Y (2018) Post-disaster infrastructure recovery: Prediction of recovery
883 rate using historical data. Reliability Engineering & System Safety 169:209–223
884 Barrientos F, Sainz G (2012) Interpretable knowledge extraction from emergency call data
885 based on fuzzy unsupervised decision tree. Knowledge-Based Systems 25:77–87
886 Barrington L, Ghosh S, Greene M, Har-Noy S, Berger J, Gill S, Lin AYM, Huyck C (2011)
887 Crowdsourcing earthquake damage assessment using remote sensing imagery. Annals of
888 Geophysics 54:680–687, DOI: 10.4401/ag-5324
889 Barthel F, Neumayer E (2012) A trend analysis of normalized insured damage from natural
890 disasters. Climate Change 113:215–237
891 Bartoli G, Fantacci R, Gei F, Marabissi D, Micciullo L (2013) A novel emergency manage-
892 ment platform for smart public safety. International Journal of Communication Systems
893 28:928–943
894 Basu M, Shandilya A, Khosla P, Ghosh K, Ghosh S (2019a) Extracting resource needs and
895 availabilities from microblogs for aiding post-disaster relief operations. IEEE Transac-
896 tions on Computational Social Systems 6:604–618
897 Basu S, Roy S, Dasbit S (2019b) A post-disaster demand forecasting system using principal
898 component regression analysis and case-based reasoning over smartphone-based DTN.
899 IEEE Transactions on Engineering Management 66:224–239
900 Bawono AS, Ali MI, Kusumadewi S, Ramli NI (2020) Methodological study to classification
901 of damage state immediately subsequent to the Banjarnegara Indonesia Earthquake on
902 2018. IOP Conference Series: Materials Science and Engineering 712:012032
903 Bayerlein H, Kerret PD, Gesbert D (2018) Trajectory optimization for autonomous flying
904 base station via reinforcement learning. In: Proceedings of the 2018 IEEE 19th Interna-
905 tional Workshop on Signal Processing Advances in Wireless Communications (SPAWC),
906 IEEE, DOI: 10.1109/SPAWC.2018.8445768
907 Bejiga M, Zeggada A, Nouffidj A, Melgani F (2017) A convolutional neural network approach
908 for assisting avalanche search and rescue operations with UAV imagery. Remote Sensing
909 9:100
910 Bellaire S, Herwijnen A, Mitterer C, Schweizer J (2017) On forecasting wet-snow avalanche
911 activity using simulated snow cover data. Cold Regions Science and Technology 144:28–
912 38
913 Bengtsson L, Lu X, Thorson A, Garfield R, von Schreeb J (2011) Improved response to
914 disasters and outbreaks by tracking population movements with mobile phone network
 Applications of artificial intelligence for disaster management 29

915 data: A post-earthquake geospatial study in Haiti. PLoS Medicine 8:e1001083

916 Benı́tez MC, Ramı́rez J, Segura JC, Ibáñez JM, Almendros J, Garcı́a-Yeguas A, Cortés G
917 (2007) Continuous HMM-based seismic-event classification at Deception Island, Antarc-
918 tica. IEEE Transactions on Geoscience and Remote Sensing 45:138–146
919 Berawi MA, Leviakangas P, Muhammad F, Sari M, Gunawan, Suryanegara YAYM (2019)
920 Optimizing search and rescue personnel allocation in disaster emergency response using
921 fuzzy logic. International Journal of Technology 10:1416–1426
922 Berkhahn S, Fuchs L, Neuweiler I (2019) An ensemble neural network model for real-time
923 prediction of urban floods. Journal of Hydrology 575:743–754
924 Besaleva LI, Weaver AC (2013) CrowdHelp: A crowdsourcing application for improving dis-
925 aster management. In: Proceedings of the 2013 IEEE Global Humanitarian Technology
926 Conference, IEEE, DOI: 10.1109/GHTC.2013.6713678
927 Bevington JS, Eguchi RT, Gill S, Ghosh S, Huyck CK (2015) A comprehensive analysis
928 of building damage in the 2010 Haiti earthquake using high-resolution imagery and
929 crowdsourcings. In: Lippitt C, Stow D, Coulter L (eds) Time-sensitive Remote Sensing,
930 Springer, New York, NY, pp 131–145
931 Bhavaraju SKT, Beyney C, Nicholson C (2019) Quantitative analysis of social media sensi-
932 tivity to natural disasters. International Journal of Disaster Risk Reduction 39:101251
933 Biswas R, Friedrich C, Rao YN, Chakraborty A (2019) Deep reinforcement learning for
934 autonomous inspection system. In: Purdue Undergraduate Research Conference, p 29
935 Boccardo P, Tonolo FG (2014) Remote sensing role in emergency mapping for disaster
936 response. Engineering Geology for Society and Territory 5:17–24
937 Bocchini P, Frangopol DM (2012a) Optimal resilience- and cost-based postdisaster interven-
938 tion prioritization for bridges along a highway segment. Journal of Bridge Engineering
939 17:117–129, DOI: 10.1061/(asce)be.1943-5592.0000201
940 Bocchini P, Frangopol DM (2012b) Restoration of bridge networks after an earth-
941 quake: Multicriteria intervention optimization. Earthquake Spectra 28:426–455, DOI:
942 10.1193/1.4000019
943 Boyd D, Crawford K (2012) Critical questions for big data. Journal of Information, Com-
944 munication & Society 15:662–679
945 Bromley E, Eisenman DP, Magana A, Williams M, Kim B, McCreary M, Chandra A,
946 Wells KB (2017) How do communities use a participatory public health approach to
947 build resilience? the Los Angeles county community disaster resilience project. In-
948 ternational Journal of Environmental Research and Public Health 14(10):1267, DOI:
949 10.3390/ijerph14101267
950 Bunker D, Levine L, Woody C (2015) Repertoires of collaboration for common operating
951 pictures of disasters and extreme events. Information Systems Frontier 17:51–65
952 Burris JW, Shrestha R, Gautam B, Bista B (2015) Machine learning for the activation
953 of contraflow during hurricane evacuation. In: The 2015 IEEE Global Humanitarian
954 Technology Conference (GHTC), IEEE, DOI: 10.1109/GHTC.2015.7343981
955 Butler D (2013) Crowdsourcing goes mainstream in typhoon response. Nature DOI:
956 10.1038/nature.2013.14186
957 Canon MJ, Satuito A, Sy C (2018) Determining disaster risk management priorities
958 through a neural network-based text classifier. In: Proceedings of the 2018 Interna-
959 tional Symposium on Computer, Consumer and Control (IS3C), IEEE, pp 237–241,
960 DOI: 10.1109/IS3C.2018.00067
961 Caragea C, Squicciarini A, Stehle S, Neppalli K, Tapia A (2014) Mapping moods: Geo-
962 mapped sentiment analysis during Hurricane Sandy. In: Proceedings of the 11th In-
963 ternational Conference on Information Systems for Crisis Response and Management
964 (ISCRAM 2014), ISCRAM, pp 642–651
965 Careem M, Silva CD, Silva RD, Raschid L, Weerawarana S (2006) Sahana: Overview of a
966 disaster management system. In: Proceedings of the 2006 International Conference on
967 Information and Automation, IEEE, pp 361–366, DOI: 10.1109/ICINFA.2006.374152
968 Casagli N, Frodella W, Morelli S, Tofani V, Ciampalini A, Intrieri E, Raspini F, Rossi G,
969 Tanteri L, Lu P (2017) Spaceborne, UAV and ground-based remote sensing techniques
970 for landslide mapping, monitoring and early warning. Geoenvironmental Disasters 4,
971 article number: 9
 30 Wenjuan Sun et al.

972 Castellanos CL, Marti JR, Sarkaria S (2018) Distributed reinforcement learning framework
973 for resource allocation in disaster response. In: 2018 IEEE Global Humanitarian Tech-
974 nology Conference (GHTC), IEEE, DOI: 10.1109/GHTC.2018.8601911
975 Cauffman SA, Dillard MK, Helgeson JF (2018) Implementation of the NIST Community
976 Resilience Planning Guide for Buildings and Infrastructure Systems. Tech. Rep. NISTIR
977 8231, National Institute of Standards and Technology, URL https://nvlpubs.nist.gov/
978 nistpubs/ir/2018/NIST.IR.8231.pdf
979 Cavalcante IM, Frazzon EM, Forcellini FA, Ivanov D (2019) A supervised machine learning
980 approach to data-driven simulation of resilient supplier selection in digital manufactur-
981 ing. International Journal of Information Management 49:86–97
982 Cavalieri F, Franchin P, Cortés JAMB, Tesfamariam S (2014) Models for seismic vulnera-
983 bility analysis of power networks: Comparative assessment. Journal of Computer-Aided
984 Civil and Infrastructure Engineering 29:590–607
985 Cerrai D, Wanik DW, Bhuiyan MAE, Zhang X, Yang J, Frediani MEB, Anagnostou EN
986 (2019) Predicting storm outages through new representations of weather and vegetation.
987 IEEE Access 7:29639–29654
988 Cervone G, Sava E, Huang Q, Schnebele E, Harrison J, Waters N (2016) Using Twitter for
989 tasking remote-sensing data collection and damage assessment: 2013 Boulder flood case
990 study. International Journal of Remote Sensing 37:100–124
991 Cha YJ, Choi W, Büyüközturk O (2017) Deep learning-based crack damage detection using
992 convolutional neural networks. Computer-Aided Civil and Infrastructure Engineering
993 32:361–378
994 Cha YJ, Buyukozturk O (2015) Structural damage detection using modal strain energy
995 and hybrid multi-objective optimization. Computer-Aided Civil and Infrastructure En-
996 gineering 30:347–358
997 Chae J, Thom D, Jang Y, Kim S, Ertl T, Ebert D (2014) Public behavior response analysis
998 in disaster events utilizing visual analytics of microblog data. Computers & Graphics
999 38:51–60, DOI: 10.1016/j.cag.2013.10.008
1000 Chang FJ, Chen PA, Lu YR, Huang E, Chang KY (2014) Real-time multi-step-ahead wa-
1001 ter level forecasting by recurrent neural networks for urban flood control. Journal of
1002 Hydrology 517:836–846
1003 Chang LC, Shen HY, Wang YF, Huang JY, Lin YT (2010) Clustering-based hybrid inunda-
1004 tion model for forecasting flood inundation depths. Journal of Hydrology 385:257–268
1005 Chang SE, Yip JZK, Conger T, Oulahen G, Marteleira M (2018) Community vulnerability
1006 to coastal hazards: Developing a typology for disaster risk reduction. Applied Geography
1007 91:81–88
1008 Chang TC, Chien YH (2007) The application of genetic algorithm in debris flows prediction.
1009 Environmental Geology 53:339–347
1010 Chaudhuri N, Bose I (2020) Exploring the role of deep neural networks for post-disaster
1011 decision support. decision support systems. Decision Support Systems 130:113234, DOI:
1012 10.1016/j.dss.2019.113234
1013 Chen LH, Hong YT (2012) Regional Taiwan rainfall frequency analysis using principal
1014 component analysis, self-organizing maps and L-moments. Hydrology Research 43:275–
1015 285
1016 Chen PA, Chang LC, Chang FJ (2013) Reinforced recurrent neural networks for multi-step-
1017 ahead flood forecasts. Journal of Hydrology 497:71–79
1018 Chen R, Sharman R, Chakravarti N, Rao HR, Upadhyaya SJ (2008a) Emergency response
1019 information system interoperability: development of chemical incident response data
1020 model. Journal of the Association for Information Systems 9:200–230
1021 Chen R, Sharman R, Rao HR, Upadhyaya SJ (2008b) Coordination in emergency response
1022 management. Communications of the ACM 51:66–73
1023 Chen S, Wang W (2009) Decision tree learning for freeway automatic incident detection.
1024 Expert Systems with Applications 36:4101–4105
1025 Chen S, Mao J, Li G, Ma C, Cao Y (2020) Uncovering sentiment and retweet patterns of
1026 disaster-related tweets from a spatiotemporal perspective a case study of Hurricane
1027 Harvey. Telematics and Informatics 47:101326
1028 Chen W, Cutter SL, Emrich CT, Shi P (2014a) Measuring social vulnerability to natural
1029 hazards in the Yangtze River Delta region, China. International Journal of Disaster Risk
 Applications of artificial intelligence for disaster management 31

1030 Science 4:169–181

1031 Chen W, Peng J, Hong H, Shahabi H, Pradhan B, Liu J, Zhu AX, Pei X, Duan Z (2018)
1032 Landslide susceptibility modelling using GIS-based machine learning techniques for
1033 Chongren County, Jiangxi Province, China. Science of the Total Environment 626:1121–
1034 1135
1035 Chen W, Yan X, Zhao Z, Hong H, Bui DT, Pradhan B (2019) Spatial prediction of land-
1036 slide susceptibility using data mining-based kernel logistic regression, naı̈ve Bayes and
1037 RBFNetwork models for the Long County area (China). Bulletin of Engineering Geology
1038 and the Environment 78:247–266
1039 Chen WT, Huang YH (2006) Approximately predicting the cost and duration of school
1040 reconstruction projects in Taiwan. Construction Management and Economics 24:1231–
1041 1239
1042 Chen Y, Niu Z, Bai J, Wang Y (2014b) Seismic vulnerability assessment of water supply
1043 network in Tianjin, China. Frontiers of Environmental Science & Engineering 8:767–775
1044 Cheng G, Guo L, Zhao T, Han J, Li H, Fang J (2013) Automatic landslide detection from
1045 remote-sensing imagery using a scene classification method based on BoVW and pLSA.
1046 International Journal of Remote Sensing 34:45–59
1047 Cheng L, Zhang J (2020) Is tourism development a catalyst of economic recovery following
1048 natural disaster? an analysis of economic resilience and spatial variability. Current Issues
1049 in Tourism DOI: 10.1080/13683500.2019.1711029
1050 Cheng MY, Hoang ND (2014) Slope collapse prediction using Bayesian framework with K-
1051 nearest neighbor density estimation: Case study in Taiwan. Journal of Computing in
1052 Civil Engineering 30:04014116
1053 Cheng X, Zhang R, Zhou J, Xu W (2017) DeepTransport: Learning spatial-temporal de-
1054 pendency for traffic condition forecasting. URL https://arxiv.org/abs/1709.09585
1055 Choi J, Yeum C, Dyke S, Jahanshahi M (2018) Computer-aided approach for rapid post-
1056 event visual evaluation of a building façade. Sensors 18:3017
1057 Chou JS, Thedja JPP (2016) Metaheuristic optimization within machine learning-based
1058 classification system for early warnings related to geotechnical problems. Automation
1059 in Construction 68:65–80
1060 Choubin B, Borji M, Mosavi A, Sajedi-Hosseini F, Singh VP, Shamshirband S (2019) Snow
1061 avalanche hazard prediction using machine learning methods. Journal of Hydrology
1062 577:123929
1063 Chung K, Park RC (2016) P2P cloud network services for IoT based disaster situations
1064 information. Peer-to-Peer Networking and Applications 9:566–577
1065 Conner AJ, Shao Y, Campbell JB (2016) Detection of urban damage using remote sensing
1066 and machine learning algorithms: Revisiting the 2010 Haiti earthquake. Remote Sensing
1067 8:868
1068 Cortez B, Carrera B, Kim YJ, Jung JY (2018) An architecture for emergency event predic-
1069 tion using LSTM recurrent neural networks. Expert Systems with Applications 97:315–
1070 324
1071 Crawford K, Finn M (2015) The limits of crisis data: analytical and ethical challenges of
1072 using social and mobile data to understand disasters. GeoJournal 80:491–502
1073 Crawford PS, Al-Zarrad MA, Graettinger AJ, Hainen AM, Back E, Powell L (2018) Rapid
1074 disaster data dissemination and vulnerability assessment through synthesis of a web-
1075 based extreme event viewer and deep learning. Advances in Civil Engineering Article
1076 ID 725156, DOI: 10.1155/2018/7258156
1077 Cresci S, Tesconi M, Cimino A, Dell’Orletta F (2015) A linguistically-driven approach to
1078 cross-event damage assessment of natural disasters from social media messages. In: Pro-
1079 ceedings of the 24th International Conference on World Wide Web, ACM, pp 1195–1200
1080 Curtis A, Mills JW, Kennedy B, Fotheringham S, McCarthy T (2007) Understanding the
1081 geography of post-traumatic stress: An academic justification for using a spatial video
1082 acquisition system in the response to Hurricane Katrina. Journal of Contingencies &
1083 Crisis Management 15:208–219
1084 Curtis A, Duval-Diop D, Novak J (2010) Identifying spatial patterns of recovery and aban-
1085 donment in the post-Katrina Holy Cross neighborhood of New Orleans. Cartography
1086 and Geographic Information Science 37:45–56
 32 Wenjuan Sun et al.

1087 Dahl GE, Yu D, Deng L, Acero A (2012) Context-dependent pre-trained deep neural net-
1088 works for large-vocabulary speech recognition. IEEE Transactions on Audio, Speech,
1089 and Language Processing 20:30–42
1090 D’Amico DF, Quiring SM, Maderia CM, McRoberts DB (2019) Improving the hurricane
1091 outage prediction model by including tree species. Climate Risk Management 25:100193
1092 Datt G, Bhatt AK, Kumar S (2015) Disaster management information system framework
1093 using feed forward back propagation neural network. International Journal of Advanced
1094 Research in Computer and Communication Engineering 4:510–514
1095 de Morsier F, Tuia D, Borgeaud M, Gass V, Thiran JP (2013) Semi-supervised novelty
1096 detection using SVM entire solution path. IEEE Transactions on Geoscience and Remote
1097 Sensing 51:1939–1950
1098 Dekanová M, Ducho F, Dekan M, Kyzek F, Biskupič M (2018) Avalanche forecasting using
1099 neural network. In: Proceedings of the 2018 ELEKTRO, IEEE, DOI: 10.1109/ELEK-
1100 TRO.2018.8398359
1101 Deng L, Yu D (2014) Deep learning: Methods and applications. Foundations and Trends in
1102 Signal Processing 7:199–200
1103 Deryugina T (2017) The fiscal cost of hurricanes: Disaster aid versus social insurance. Amer-
1104 ican Economic Journal: Economic Policy 9:168–198
1105 Devault JE (2000) Robotic system for underwater inspection of bridge piers. IEEE Instru-
1106 mentation & Measurement Magazine 3:32–37
1107 DeVries PMR, Vigas F, Wattenberg M, Meade BJ (2018) Deep learning of aftershock pat-
1108 terns following large earthquakes. Nature 560:632–634
1109 DHS (2010) DHS risk lexicon. URL https://www.dhs.gov/xlibrary/assets/
1110 dhs-risk-lexicon-2010.pdf
1111 Djordjevich DD, Xavier PG, Bernard ML, Whetzel JH, Glickman MR, Verzi SJ (2008)
1112 Preparing for the aftermath: Using emotional agents in game-based training for disaster
1113 response. In: The 2008 IEEE Symposium On Computational Intelligence and Games,
1114 IEEE, pp 266–275
1115 DOE (2014) Lantern live. URL https://github.com/GSA/digitalgov.gov/blob/master/
1116 content/posts/2014/12/2014-12-04-find-fuel-during-disasters-with-lantern-live-app.
1117 md
1118 Dogaru DI, Dumitrache I (2019) Cyber attack of a power grid analysis using a deep neural
1119 networks approach. Journal of Control Engineering and Applied Informatics 21:42–50
1120 Dong L, Wei F, Tan C, Tang D, Zhou M, Xu K (2014) Adaptive recursive neural network for
1121 target-dependent Twitter sentiment classification. In: Proceedings of the 52nd Annual
1122 Meeting of the Association for Computational Linguistics (Short Papers), Association
1123 for Computational Linguistics, pp 49–54
1124 Dou J, Yamagishi H, Pourghasemi HR, Yunus AP, Song X, Xu Y, Zhu Z (2015) An in-
1125 tegrated artificial neural network model for the landslide susceptibility assessment of
1126 Osado Island, Japan. Natural Hazards 78:1749–1776
1127 Dou M, Chen J, Chen D, Chen X, Deng Z, Zhang X, Xua K, Wang J (2014) Modeling and
1128 simulation for natural disaster contingency planning driven by high-resolution remote
1129 sensing images. Future Generation Computer Systems 37:367–377
1130 Duffey RB (2019) Power restoration prediction following extreme events and disasters. In-
1131 ternational Journal of Disaster Risk Science 10:134–148
1132 Duncan A, Chen AS, Keedwell E, Djordjević S, Savić DA (2013) RAPIDS: early warn-
1133 ing system for urban flooding and water quality hazards. In: Proceedings of Ma-
1134 chine Learning in Water Systems Symposium: part of AISB Annual Convention 2013,
1135 http://hdl.handle.net/10871/16090
1136 Dutta I, Dutta S, Raahemi B (2017) Detecting financial restatements using data mining
1137 techniques. Expert Systems with Applications 90:374–393
1138 Eckle M, Herfort B, Yan Y, Kuo CL, Zipf A (2017) Towards using volunteered geographic
1139 information to monitor post-disaster recovery in tourist destinations. In: Proceedings of
1140 the 14th ISCRAM Conference, pp 1008–1019
1141 Eguchi RT, Goltz JD, Taylor CE, Chang SE, Flores PJ, Johnson LA, Seligson HA, Blais
1142 NC (1998) Direct economic lossed in the Northridge Earthquake: a three-year post-event
1143 perspective. Earthquake Spectra 14:245–264
 Applications of artificial intelligence for disaster management 33

1144 Eguchi RT, Huyck CK, Ghosh S, Adams BJ (2008) The application of remote sensing
1145 technologies for disaster management. In: The 14th World Conference on Earthquake
1146 Engineering, Paper ID. K004, https://www.iitk.ac.in/nicee/wcee/article/14 K004.pdf
1147 Eicken H, Jones J, Meyer F, Mahoney A, Druckenmiller ML, Rohith M, Kambhamettu
1148 C (2011) Environmental security in Arctic ice-covered seas: From strategy to tactics
1149 of hazard identification and emergency response. Marine Technology Society Journal
1150 45:37–48
1151 Eid MS, El-adaway IH (2017a) Integrating the social vulnerability of host communities and
1152 the objective functions of associated stakeholders during disaster recovery processes
1153 using agent-based modeling. Journal of Computing in Civil Engineering 31:04017030
1154 Eid MS, El-adaway IH (2017b) Sustainable disaster recovery: Multiagent-based model for
1155 integrating environmental vulnerability into decision-making processes of the associated
1156 stakeholders. Journal of Urban Planning and Development 143:04016022
1157 Eisenman DP, Chandra A, Fogleman S, Magana A, Hendricks A, Wells KB, Williams M,
1158 Tang J, Plough A (2014) The Los Angeles county community disaster resilience project
1159 a community-level, public health initiative to build community disaster resilience. In-
1160 ternational Journal of Environmental Research and Public Health 11:8475–8490
1161 Ellenberg A, Branco L, Krick A, Bartoli I, Kontsos A (2015) Use of unmanned aerial vehicle
1162 for quantitative infrastructure evaluation. Journal of Infrastructure Systems 21:04014054
1163 Ellingwood BR, Cutler H, Gardoni P, Peacock WG, van de Lindt JW, Wang N (2016) The
1164 Centerville virtual community: a fully integrated decision model of interacting physical
1165 and social infrastructure systems. Sustainable and Resilient Infrastructure 1:95–107
1166 Elsayed M, Erol-Kantarc M (2018) Deep Q-learning for low-latency tactile applications: Mi-
1167 crogrid communications. In: Proceedings of the 2018 IEEE International Conference on
1168 Communications, Control, and Computing Technologies for Smart Grids (SmartGrid-
1169 Comm), IEEE, DOI: 10.1109/SmartGridComm.2018.8587476
1170 EPA (2016) Incident waste decision support tool (I-WASTE DST). URL http://www2.
1171 ergweb.com/bdrtool/login.asp
1172 Eskandarpour R, Khodaei A (2017) Machine learning based power grid outage prediction
1173 in response to extreme events. IEEE Transactions on Power Systems 32:3315–3316
1174 ESRI (2016) Disaster response program. URL https://www.esri.com/en-us/
1175 disaster-response/overview
1176 Ettinger S, Mounaud L, Magill C, Yao-Lafourcade AF, Thouret JC, Manville V, Negulescu
1177 C, Zuccaro G, Gregorio DD, Nardone S, Uchuchoque JAL, Arguedas A, Macedo L,
1178 Llerena NM (2016) Building vulnerability to hydro-geomorphic hazards: Estimating
1179 damage probability from qualitative vulnerability assessment using logistic regression.
1180 Journal of Hydrology 541:563–581
1181 EU-CIRCLE (2019) A pan-European framework for strengthening critical infrastructure
1182 resilience to climate change. URL https://www.eu-circle.eu/eu-funded-projects/
1183 Fallahian M, Khoshnoudian F, Meruane V (2018) Ensemble classification method for struc-
1184 tural damage assessment under varying temperature. Structural Health Monitoring
1185 17:747–762
1186 Fan C, Zhang C, Yahja A, Mostafavi A (2019) Disaster City Digital Twin: A vision for inte-
1187 grating artificial and human intelligence for disaster management. International Journal
1188 of Information Management DOI: 10.1016/j.ijinfomgt.2019.102049
1189 FEMA (2018) Hazard mitigation grant program. https://www.fema.gov/
1190 hazard-mitigation-grant-program
1191 FEMA (2019) Federal insurance & mitigation administration national flood insurance pro-
1192 gram (FIMA NFIP) redacted claims dataset. https://www.fema.gov/media-library/
1193 assets/documents/180374
1194 Feng Q, Liu J, Gong J (2019) Urban flood mapping based on unmanned aerial vehicle remote
1195 sensing and random forest classifiera case of Yuyao, China. Water 7:1437–1455
1196 Fernandez P, Mourato S, Moreira M, Pereira L (2016) A new approach for computing a
1197 flood vulnerability index using cluster analysis. Physics and Chemistry of the Earth,
1198 Parts A/B/C 94:47–55
1199 Fiedrich F, Gehbauer F, Rickers U (2000) Optimized resource allocation for emergency
1200 response after earthquake disasters. Safety Science 35:41–57
 34 Wenjuan Sun et al.

1201 Fohringer J, Dransch D, Kreibich H, Schröter K (2015) Social media as an information

1202 source for rapid flood inundation mapping. Natural Hazards and Earth System Sciences
1203 15:2725–2738
1204 Foresti GL, Farinosi M, Vernier M (2015) Situational awareness in smart environments:
1205 socio-mobile and sensor data fusion for emergency response to disasters. Journal of
1206 Ambient Intelligence and Humanized Computing 6:239–257
1207 Fotovatikhah F, Herrera M, Shamshirband S, Chaue K, Ardabili SF, Piran MJ (2018) Survey
1208 of computational intelligence as basis to big flood management: challenges, research
1209 directions and future work. Engineering Applications of Computational Fluid Mechanics
1210 12:411–437
1211 Galatzer-Levy IR, Karstoft KI, Statnikov A (2014) Quantitative forecasting of PTSD from
1212 early trauma responses: A machine learning application. Journal of Psychiatric Research
1213 59:68–76
1214 Gama M, Santos BF, Scaparra MP (2016) A multi-period shelter location-allocation model
1215 with evacuation orders for flood disasters. EURO Journal on Computational Optimiza-
1216 tion 4:299–323
1217 Ganesan P, Sathish BS, Sajiv G (2016) A comparative approach of identification and segmen-
1218 tation of forest fire region in high resolution satellite images. In: 2016 World Conference
1219 on Futuristic Trends in Research and Innovation for Social Welfare (Startup Conclave),
1220 IEEE, DOI: 10.1109/STARTUP.2016.7583959
1221 Gao X, Nayeem MK, Hezam IM (2019) A robust two-stage transit-based evacuation model
1222 for large-scale disaster response. Measurement 145:713–723
1223 Gao Y, Chen YX, Ding YS, Tang BY (2006) Immune genetic algorithm based on network
1224 model for flood disaster evaluation. Journal of Natural Disaster 15:110–114
1225 Gauthier F, Germain D, Huétu B (2017) Logistic models as a forecasting tool for snow
1226 avalanches in a cold maritime climate: northern Gaspésie, Québec, Canada. Natural
1227 Hazards 89:201–232
1228 Geiß C, Taubenböck H, Tyagunov S, Tisch A, Post J, Lakes T (2014) Assessment of seismic
1229 building vulnerability from space. Earthquake Spectra 30:1553–1583
1230 Gemici MC, Savena A (2014) Learning haptic representation for manipulating deformable
1231 food objects. In: Proceedings of the 2014 IEEE/RSJ International Conference on Intel-
1232 ligent Robots and Systems, IEEE, pp 638–645, DOI: 10.1109/IROS.2014.6942626
1233 Geng B (2019) Traffic prediction and transmission scheduling of artificial intelligence-driven
1234 cognitive wireless networks. International Journal of Computers and Applications DOI:
1235 10.1080/1206212X.2019.1706812
1236 GeoPlatform (2016) Geoplatform.gov. URL https://www.Disasters.GeoPlatform.gov/
1237 German S, Jeon JS, Zhu Z, Bearman C, Brilakis I, DesRoches R, Lowes L (2013) Machine
1238 vision-enhanced postearthquake inspection. Journal of Computing in Civil Engineering
1239 27:622–634
1240 Ghaffarian S, Kerle N, Pasolli E, Arsanjani JJ (2019) Post-disaster building database up-
1241 dating using automated deep learning: An integration of pre-disaster OpenStreetMap
1242 and multi-temporal satellite data. Remote Sensing 11:2427
1243 Ghosh J, Padgett J, nas Osorio LD (2013) Surrogate modeling and failure surface visual-
1244 ization for efficient seismic vulnerability assessment of highway bridges. Probabilistic
1245 Engineering Mechanics 34:189–199
1246 Ghosh S, Huyck CK, Greene M, Gill SP, Bevington J, Svekla W, DesRoches R, Eguchi
1247 RT (2011) Crowdsourcing for rapid damage assessment: The global earth observation
1248 catastrophe assessment network (GEO-CAN). Earthquake Spectra 27:S179–S198
1249 Ghosh T, Krishnamurti TN (2018) Improvements in hurricane intensity forecasts from a
1250 multimodel superensemble utilizing a generalized neural network technique. Weather
1251 and Forecasting 33(3):873–885, DOI: 10.1175/WAF-D-17-0006.1
1252 Giffard-Roisin S, Yang M, Charpiat G, Kégl B, Monteleoni C (2018) Deep learning for
1253 hurricane track forecasting from aligned spatio-temporal climate datasets. In: Workshop
1254 on Modeling and Decision-Making in the Spatiotemporal Domain, the 32nd Conference
1255 on Neural Information Processing Systems (NIPS 2018), https://openreview.net/pdf?
1256 id=rkMdBSdRKm
1257 Gilmour PM (2019) The application of photography in investigating fraud. The Imaging
1258 Science Journal 67(4):215–223, DOI: 10.1080/13682199.2019.1600254
 Applications of artificial intelligence for disaster management 35

1259 Giusti A, Guzzi J, Cireşan DC, He FL, Rodrguez JP, Fontana F, Faessler M, Forster C,
1260 Schmidhuber J, Caro GD, Scaramuzza D, Gambardella LM (2015) A machine learning
1261 approach to visual perception of forest trails for mobile robots. IEEE Robotics and
1262 Automation Letters 1:661–667, DOI: 10.1109/LRA.2015.2509024
1263 Goetz JN, Brenning A, Petschko H, Leopold P (2015) Evaluating machine learning and sta-
1264 tistical prediction techniques for landslide susceptibility modeling. Computers & Geo-
1265 sciences 81:1–11
1266 Gomes GF, Almeida FA, Junqueira DM, Cunha Jr SS, Ancelotti Jr AC (2019) Optimized
1267 damage identification in CFRP plates by reduced mode shapes and GA-ANN methods.
1268 Engineering Structures 181:111–123
1269 Gomez C, Purdie H (2016) UAV-based photogrammetry and geocomputing for hazards and
1270 disaster risk monitoring – a review. Geoenvironmental Disasters 3:23
1271 Gong Q, Li L, Tognin S, Wu Q, Pettersson-Yeo W, Lui S, Huang X, Marquand AF, Mechelli
1272 A (2013) Using structural neuroanatomy to identify trauma survivors with and without
1273 post-traumatic stress disorder at the individual level. Psychological Medicine 44:195–203
1274 Gonzalez MC, Hidalgo CA, Barabasi AL (2009) Understanding individual human mobility
1275 patterns. Nature 453:779–782
1276 Goodchild MF, Glennon JA (2010) Crowdsourcing geographic information for disaster re-
1277 sponse: a research frontier. International Journal of Digital Earth 3:231–241, DOI:
1278 10.1080/17538941003759255
1279 Gopnarayan A, Deshpande S (2019) Tweets analysis for disaster management: Preparedness,
1280 emergency response, impact, and recovery. In: International Conference on Innovative
1281 Data Communication Technologies and Application, Springer, pp 760–764
1282 Grasic V, Kos A, Mileva-Boshkoska B (2018) Classification of incoming calls for the
1283 capital city of Solvenia smart city 112 public safety system using open Internet of
1284 Things data. International Journal of Distributed Sensor Networks 14(9):1–12, DOI:
1285 10.1177/1550147718801703
1286 Graves A, rahman Mohamed A, Hinton G (2013) Speech recognition with deep recurrent
1287 neural networks. URL https://arxiv.org/abs/1303.5778
1288 Greifeneder R, Bless H, Pham M (2011) When do people rely on affective and cognitive
1289 feelings in judgment?: A review. Personality and Social Psychology Review 15:107–141
1290 Guha S, Rastogi R, Shim K (1998) CURE: an efficient clustering algorithm for large
1291 databases. In: Proceeedings of the 1998 ACM SIGMOD International Conference on
1292 Management of Data, ACM, pp 73–84
1293 Günay O, Taşdemir K, Töreyin BU, Çetin AE (2010) Fire detection in video using LMS
1294 based active learning. Fire Technology 46:551–577
1295 Gunning D, Stefik M, Choi J, Miller T, Stumpf S, Yang GZ (2019) XAI–explainable artificial
1296 intelligence. Science Robotics 4:eaay7120
1297 Guo J, Wu J, Guo J, Jiang Z (2018) A damage identification approach for offshore jacket
1298 platforms using partial modal results and artificial neural networks. Applied Sciences
1299 8:2173
1300 Guo J, Huo Y, Shi X, Wu J, Yu P, Feng L, Li W (2019) 3D aerial vehicle base sta-
1301 tion (UAV-BS) position planning based on deep Q-learning for capacity enhancement
1302 of users with different QoS requirements. In: Proceedings of the 15th International
1303 Wireless Communications & Mobile Computing Conference (IWCMC), IEEE, DOI:
1304 10.1109/IWCMC.2019.8766625
1305 Guo X, Yuan Z, Tian B (2009) Supplier selection based on hierarchical potential support
1306 vector machine. Expert Systems with Applications 36:6978–6985
1307 Gupta R, Hosfelt R, Sajeev S, Patel N, Goodman B, Doshi J, Heim E, Choset H, Gaston
1308 M (2019) xBD: A dataset for assessing building damage from satellite imagery. In:
1309 Proceedings of the CVPR Workshops 2019, arXiv: 1911.09296v1
1310 Hackl J, Adey BT, Lethanh N (2018) Determination of nearoptimal restoration programs
1311 for transportation networks following natural hazard events using simulated annealing.
1312 Computer-Aided Civil and Infrastructure Engineering 33:618–637
1313 Han W, Tian Z, Huang Z, Huang D, Jia Y (2019) Quantitative assessment of wireless
1314 connected intelligent robot swarms network security situation. IEEE Access 7:134293 –
1315 134300
 36 Wenjuan Sun et al.

1316 Haraoka T, Ojima T, Murata C, Hayasaka S (2012) Factors influencing collaborative activ-
1317 ities between non-professional disaster volunteers and victims of earthquake disasters.
1318 PLoS One 7:e47203
1319 Hartawan DR, Purboyo TW, Setianingsih C (2019) Disaster victims detection system using
1320 convolutional neural network (CNN) method. In: 2019 IEEE International Conference
1321 on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT), IEEE,
1322 pp 105–111, DOI: 10.1109/ICIAICT.2019.8784782
1323 Hasan S, Ukkusuri SV (2014) Urban activity pattern classification using topic models
1324 from online geo-location data. Transportation Research Part C: Emerging Technolo-
1325 gies 44:363–381, DOI: 10.1016/j.trc.2014.04.003
1326 Hashemipour M, Stuban SMF, Dever JR (2017) A community-based disaster coordination
1327 framework for effective disaster preparedness and response. Australian Journal of Emer-
1328 gency Management 32:41–46
1329 Heck M, Hammer C, van Herwijnen A, Schweizer J, Fäh D (2010) Automatic detection
1330 of snow avalanches in continuous seismic data using hidden Markov models. Natural
1331 Hazards and Earth System Sciences 18:383–396
1332 Hernández E, Sanchez-Anguix V, Julian V, Palanca J, Duque N (2016) Rainfall prediction:
1333 a deep learning approach. In: Proceedings of the International Conference on Hybrid
1334 Artificial Intelligence Systems (HASI2016), pp 151–162
1335 Hernandez-Suarez A, Sanchez-Perez G, Toscano-Medina K, Perez-Meana H, Portillo-Portillo
1336 J, Sanchez V, Villalba LJG (2019) Using Twitter data to monitor natural disaster
1337 social dynamics: A recurrent neural network approach with word embeddings and kernel
1338 density estimation. Sensors 19:1746
1339 Heß VDC (2017) Weigh(t)ing the dimensions of social vulnerability based on a regres-
1340 sion analysis of disaster damages. Natural Hazards and Earth System Science DOI:
1341 10.5194/nhess-2017-74
1342 Higuchi H, Fujimura J, Nakamura T, Kogo K, Tsudaka K, Wada T (2014) Disaster de-
1343 tection by statistics and SVM for emergency rescue evacuation support system. In:
1344 The 43rd International Conference on Parallel Processing Workshops, IEEE, DOI:
1345 10.1109/ICPPW.2014.52
1346 Hochgraf C, Nygate J, Bazdresch M, Indelicato M, Johnson WP, Espinos R (2018) Providing
1347 first responders with real-time status of cellular networks during a disaster. In: The 2018
1348 IEEE International Symposium on Technologies for Homeland Security (HST), IEEE,
1349 DOI: 10.1109/THS.2018.8574145
1350 Hoeppe P (2016) Trends in weather related disasters consequences for insurers and society.
1351 Weather and Extreme Events 11:70–79
1352 Holgado P, Vollagrá VA, Vázques L (2017) Real-time multistep attack prediction based
1353 on hidden Markov models. IEEE Transactions on Dependable and Secure Computing
1354 17:134–147
1355 Holguı́n-Veras J, Jaller M, Wassenhove LNV, Pérez N, Wachtendorf T (2012) On the unique
1356 features of post-disaster humanitarian logistics. Journal of Operations Management
1357 30:494–506
1358 Hong L, Yin D, Guo J, Davison BD (2011) Tracking trends: incorporating term volume
1359 into temporal topic models. In: Proceedings of the 17th ACM SIGKDD international
1360 conference on Knowledge discovery and data mining, ACM, pp 484–492
1361 Hoot N, Aronsky D (2006) An early warning system for overcrowding in the emergency
1362 department. AMIA Annual Symposium Proceedings Symposium 2006:339–343, PMCID:
1363 PMC1839284
1364 Hosseini FS, Choubin B, Mosavi A, Nabipour N, Shamshirband S, Darabi H, Haghighih
1365 AT (2019) Flash-flood hazard assessment using ensembles and bayesian-based machine
1366 learning models: Application of the simulated annealing feature selection method. Sci-
1367 ence of the Total Environment 711:135161, DOI: 10.1016/j.scitotenv.2019.135161
1368 Hosseini-Moghari SM, Araghinejad S (2015) Monthly and seasonal drought forecasting using
1369 statistical neural networks. Environmental Earth Sciences 74:397–412
1370 Hou MC, Deng DJ, Wu CL (2019) Optimum aerial base station deployment for UAV net-
1371 works: A reinforcement learning approach. In: 2019 IEEE Globecom Workshops, IEEE,
1372 DOI: 10.1109/GCWkshps45667.2019.9024648
 Applications of artificial intelligence for disaster management 37

1373 Hou TC, Liu JW, Liu YW (2017) Algorithmic clustering of LiDAR point cloud data for
1374 textural damage identifications of structural elements. Measurement 108:77–90
1375 Hu D, Li S, Chen J, Kamat VR (2019) Detecting, locating, and characterizing voids in
1376 disaster rubble for search and rescue. Advanced Engineering Informatics 42:100974
1377 Huang L, Xiang L (2018) Method for meteorological early warning of precipitation-induced
1378 landslides based on deep neural network. Neural Processing Letters 48:1143–1260
1379 Huang M, Lei Y, Cheng S (2019a) Damage identification of bridge structure considering
1380 temperature variations based on particle swarm optimization - cuckoo search algorithm.
1381 Advances in Structural Engineering 22:3262–3276
1382 Huang W, Wang Y, Yi X (2017) Deep Q-learning to preserve connectivity in multi-robot sys-
1383 tems. In: Proceedings of the 9th International Conference on Signal Processing Systems
1384 (ICSPS2017), ACM, pp 45–50
1385 Huang X, Li Z, Wang C, Ning H (2019b) Identifying disaster related social media for rapid
1386 response: a visual-textual fused CNN architecture. International Journal of Digital Earth
1387 DOI: 10.1080/17538947.2019.1633425
1388 Huang X, Wang C, Li Z, Ning H (2019c) A visualtextual fused approach to automated
1389 tagging of flood-related tweets during a flood event. International Journal of Digital
1390 Earth 12:1248–1264
1391 Huang Y, Jin L, Zhao H, Huang X (2018) Fuzzy neural network and LLE algorithm for
1392 forecasting precipitation in tropical cyclones: comparisons with interpolation method
1393 by ECMWF and stepwise regression method. Natural Hazards 91:201–220
1394 Huang Z, Zhou J, Song L, Lu Y, Zhang Y (2010) Flood disaster loss comprehensive evalua-
1395 tion model based on optimization support vector machine. Expert Systems with Appli-
1396 cations 37:3810–3814
1397 Hung KC, Kalantari M, Rajabifard A (2016) Methods for assessing the credibility of vol-
1398 unteered geographic information in flood response: A case study in Brisbane, Australia.
1399 Applied Geography 68:37–47, DOI: 10.1016/j.apgeog.2016.01.005
1400 Hutchinson TC, Chen Z (2005) Optimized estimated ground truth for object-based urban
1401 damage estimation using satellite images from the 2003 Bam, Iran, Earthquake. Earth-
1402 quake Spectra 21:S239–S254
1403 Ifrim G, Shi B, Brigadir I (2014) Event detection in Twitter using aggressive filtering and
1404 hierarchical tweet clustering. In: The Second Workshop on Social News on the Web
1405 (SNOW), URL http://hdl.handle.net/10197/7546
1406 Iliadis LS (2005) A decision support system applying an integrated fuzzy model for long-term
1407 forest fire risk estimation. Environmental Modelling & Software 20:613–621
1408 Ilyas A (2014) MicroFilters: harnessing Twitter for disaster management. In: Proceedings of
1409 the IEEE Global Humanitarian Technology Conference (GHTC 2014), IEEE, pp 417–
1410 424, DOI: 10.1109/GHTC.2014.6970316
1411 Imran M, Elbassuoni S, Castillo C, Diaz F (2013) Extracting information nuggets from
1412 disaster-related messages in social media. In: The 10th International Conference on
1413 Information Systems for Crisis Response and Management (ISCRAM), https://pdfs.
1414 semanticscholar.org/ba89/77eb5d737d643f38d84d0c1476211bb88986.pdf
1415 Imran M, Castillo C, Lucas J, Meier P, Vieweg S (2014) AIDR: artificial intelligence for
1416 disaster response. In: Proceedings of the 23rd International Conference on World Wide
1417 Web, ACM, pp 159–162
1418 Ireland G, Volpi M, Petropoulos GP (2015) Examining the capability of supervised machine
1419 learning classifiers in extracting flooded areas from landsat tm imagery: A case study
1420 from a Mediterranean flood. Remote Sensing 7:3372–3399
1421 Izadi M, Mohammadzade A, Haghighattalab A (2017) A new neuro-fuzzy approach for post-
1422 earthquake road damage assessment using GA and SVM classification from Quickbird
1423 satellite images. Journal of the Indian Society of Remote Sensing 45:965–977
1424 Izumi T, Shaw R, Djalante R, Ishiwatari M, Komino T (2019) Disaster risk reduction and
1425 innovations. Progress in Disaster Science 2:100033
1426 Jaech A, Zhang B, Ostendorf M, Kirschen DS (2019) Real-time prediction of the duration
1427 of distribution system outages. IEEE Transactions on Power Systems 34:773–781
1428 Jahangiri A, Afandizadeh S, Kalantari N (2011) The optimization of traffic signal timing for
1429 emergency evacuation using the simulated annealing algorithm. Transport 26:133–140
 38 Wenjuan Sun et al.

1430 Jamali M, Nejat A, Ghosh S, Jin F, Cao G (2019) Social media data and post-disaster
1431 recovery. International Journal of Information Management 44:25–37
1432 Jayaram N, Baker JW (2010) Efficient sampling and data reduction techniques for proba-
1433 bilistic seismic lifeline risk assessment. Earthquake Engineering & Structural Dynamics
1434 39:1109–1131
1435 Jhong BC, Wang JH, Lin GF (2017) An integreated two-stage support vector machine
1436 approach to forecast inundation maps during typhoons. Journal of Hydrology 547:236–
1437 252
1438 Jia M, Zhang Z (2012) Critical mass of women on BODs, multiple identities, and corporate
1439 philanthropic disaster response: Evidence from privately owned Chinese firms. Journal
1440 of Business Ethics 118:303–317, DOI: 10.1007/s10551-012-1589-7
1441 Jiang FZ, Zhong L, Thilakarathna K, Seneviratne A, Takano K, Yamada S, Ji Y (2017)
1442 Supercharging crowd dynamics estimation in disasters via spatio-temporal deep neural
1443 network. In: The 2017 IEEE International Conference on Data Science and Advanced
1444 Analytics (DSAA), IEEE, DOI: 10.1109/DSAA.2017.11
1445 Jiang S, Friedland CJ (2016) Automatic urban debris zone extraction from post-hurricane
1446 very high-resolution satellite and aerial imagery. Journal of Geomatics, Natural Hazards
1447 and Risk 7:1–20
1448 Jones SS, Thomas A, Evans RS, Welch SJ, Haug PJ, Snow GL (2008) Forecasting daily
1449 patient volumes in the emergency department. Academic Emergency Medicine 15:159–
1450 170
1451 Joo S, Ogawa Y, Sekimoto Y (2019) Decision-making system for road-recovery considering
1452 human mobility by applying deep Q-network. In: The 2019 IEEE International Confer-
1453 ence on Big Data, IEEE, DOI: 10.1109/BigData47090.2019.9006385
1454 Kahn ME (2006) The death toll from natural disasters: The role of income, geography, and
1455 institutions. Review of Economics and Statistics 87:271–284
1456 Kameshwar S, Padgett JE (2014) Multi-hazard risk assessment of highway bridges subjected
1457 to earthquake and hurricane hazards. Engineering Sturctures 78:154–166
1458 Kamilaris A, Boldú FXP (2017) Disaster monitoring using unmanned aerial vehicles and
1459 deep learning. In: Disaster Management for Resilience and Public Safety Workshop
1460 Karamlou A, Bocchini P (2016) Sequencing algorithm with multiple-input genetic operators:
1461 Application to disaster resilience. Engineering Structures 117:591–602
1462 Karamlou A, Bocchini P, Christou V (2016) Metrics and algorithm for optimal retrofit
1463 strategy of resilient transportation networks. In: Proceedings of the Maintenance, Mon-
1464 itoring, Safety, Risk and Resilience of Bridges and Bridge Networks (IABMAS2016), pp
1465 1121–1128
1466 Karamouz M, Nazif MK, Nazif S (2014) Prediction of sea level using a hybrid data-driven
1467 model: New challenges after Hurricane Sandy. Water Quality Exposure and Health
1468 6:63–71
1469 Karstoft KI, Galatzer-Levy IR, Statnikov A, Li Z, Shalev AY (2015) Bridging a translational
1470 gap: using machine learning to improve the prediction of PTSD. BMC Psychiatry 15,
1471 article No. 30
1472 Kellermann P, Schröter K, Thieken AH, Haubrock SN, Kreibich H (2020) The object-specific
1473 flood damage database HOWAS21. Natural Hazards and Earth System Sciences DOI:
1474 10.5194/nhess-2019-420
1475 Khadr M (2016) Forecasting of meteorological drought using Hidden Markov Model (case
1476 study: The upper Blue Nile river basin, Ethiopia). Ain Shams Engineering Journal 7:47–
1477 56
1478 Khaloo A, Lattanzi D, Cunningham K, DellAndrea R, Riley M (2017) Unmanned aerial
1479 vehicle inspection of the Placer River trail bridge through image-based 3D modelling.
1480 Structure and Infrastructure Engineering 14:124–136
1481 Khan MAU, Sayem MA (2012) Understanding recovery of small enterprises from natural
1482 disaster. Environmental Hazards 12:218–239
1483 Khan SH, He X, Porikli F, Bennamoun M (2017) Forest change detection in incomplete
1484 satellite images with deep neural networks. IEEE Transactions on Geoscience and Re-
1485 mote Sensing 55:5407 – 5423
1486 Khan TA, Alam M, Kadir K, Shahid Z, Mazliham M (2018) A novel approach for the
1487 investigation of flash floods using soil flux and CO2: An implementation of MLP with less
 Applications of artificial intelligence for disaster management 39

1488 false alarm rate. In: Proceedings of the 2nd International Conference on Smart Sensors
1489 and Application (ICSSA), IEEE, pp 130–134, DOI: 10.1109/ICSSA.2018.8535606
1490 Khoshnoudian F, Talaei S, Fallahian M (2017) Structural damage detection using FRF
1491 data, 2D-PCA, artificial neural networks and imperialist competitive algorithm simul-
1492 taneously. International Journal of Structural Stability and Dynamics 17:1750073
1493 Khouj M, López C, Sarkaria S, Marti J (2011) Disaster management in real time simulation
1494 using machine learning. In: Proceedings of the 24th Canadian Conference on Electrical
1495 and Computer Engineering(CCECE), IEEE, DOI: 10.1109/CCECE.2011.6030716
1496 Kiatpanont R, Tanlamai U, Chongstitvatana P (2016) Exaction of actionable information
1497 from crowdsourced disaster data. Journal of Emergency Management 14:377–390
1498 Kim D, joo Lee H, Cho S (2008) Response modeling with support vector regression. Expert
1499 Systems with Applications 34:1102–1108
1500 Kim DW, Deo RC, Lee JS, Yeom JM (2017) Mapping heatwave vulnerability in korea.
1501 Natural Hazards 89:35–55
1502 Kim J, Sul SH, Choi JB (2018a) Development of unmanned remote smart rescue platform
1503 applying Internet of Things technology. International Journal of Distributed Sensor
1504 Networks 14, DOI: 10.1177/1550147718784482
1505 Kim JM, Woods PK, Park YJ, Kim T, Son K (2015) Predicting hurricane wind damage by
1506 claim payout based on Hurricane Ike in Texas. Geomatics, Natural Hazards and Risk
1507 76:565–585
1508 Kim JM, Wood PK, Park YJ, Son K (2016a) Estimating the Texas Windstorm Insurance
1509 Association claim payout of commercial buildings from Hurricane Ike. Natural Hazards
1510 84:405–424
1511 Kim JM, Kim T, Yu YJ, Son K (2018b) Development of a maintenance and repair cost
1512 estimation model for educational buildings using regression analysis. Journal of Asian
1513 Architecture and Building Engineering 17:307–312
1514 Kim JM, Son K, Kim YJ (2019) Assessing regional typhoon risk of disaster management by
1515 clustering typhoon paths. Environment, Development and Sustainability 21:2083–2096
1516 Kim S, Kim H, Namkoong Y (2016b) Ordinal classification of imbalanced data with applica-
1517 tion in emergency and disaster information services. IEEE Intelligent Systems 31:50–56
1518 Kim S, Lee W, Park Y, Lee H, Lee Y (2016c) Forest fire monitoring system based on
1519 aerial image. In: Proceedings of the 3rd International Conference on Information and
1520 Communication Technologies for Disaster Management (ICT-DM), IEEE, pp 1–6
1521 Kim Y, Ghorpade A, Zhao F, Pereira FC, Zegras PC, Ben-Akiva M (2018c) Activity recog-
1522 nition for a smartphone and web-based human mobility sensing system. IEEE Intelligent
1523 Systems 33:5–23
1524 Ko B, Kwak S (2012) Survey of computer vision-based natural disaster warning systems.
1525 Optical Engineering 51:070901
1526 Koch C, Paal SG, Rashidi A, Zhu Z, König M, Brilakis I (2016) Achievements and challenges
1527 in machine vision-based inspection of large concrete structures. Advances in Structural
1528 Engineering 17:303–318
1529 Kochersberger K, Kroeger K, Krawiec B, Brewer E, Weber T (2014) Postdisaster remote
1530 sensing and sampling via an autonomous helicopter. Journal of Field Robotics 31:510–
1531 521
1532 Kondaveti R, Ganz A (2009) Decision support system for resource allocation in disaster man-
1533 agement. In: The Annual International Conference of the IEEE Engineering in Medicine
1534 and Biology Society, IEEE, pp 3425–3428, DOI: 10.1109/IEMBS.2009.5332498
1535 Kong Q, Allen RM, Schreier L, Kwon YW (2016a) MyShake: A smartphone seismic network
1536 for earthquake early warning and beyond. Science Advances 2:e1501055
1537 Kong SG, Jin D, Li S, Kim H (2016b) Fast fire flame detection in surveillance video using
1538 logistic regression and temporal smoothing. Fire Safety Journal 79:37–43
1539 Kousky C, MichelKerjan E (2015) Examining flood insurance claims in the United States:
1540 Six key findings. The Journal of Risk and Insurance 84:819–850
1541 Kovordányi R, Roy C (2009) Cyclone track forecasting based on satellite images using arti-
1542 ficial neural networks. ISPRS Journal of Photogrammetry and Remote Sensing 64:513–
1543 521
1544 Krizhevsky A, Sutskever I, Hinton GE (2017) ImageNet classification with deep convolu-
1545 tional neural networks. Communications of the ACM 60:1097–1105
 40 Wenjuan Sun et al.

1546 Kryvasheyeu Y, Chen H, Obradovich N, Moro E, Hentenryck PV, Fowler J, Cebrian M

1547 (2016) Rapid assessment of disaster damage using social media activity. Science Ad-
1548 vances 2(3):e1500779, DOI: 10.1126/sciadv.1500779
1549 Kuang S, Davison BD (2017) Learning word embeddings with Chi-Square weights for health-
1550 care tweet classification. Applied Sciences 7(8), article 846
1551 Kumar A, Jiang M, Fang Y (2014) Where not to go?: Detecting road hazards using Twit-
1552 ter. In: Proceedings of the 37th International ACM SIGIR Conference on Research &
1553 Development in Information Retrieval, ACM, pp 1223–1226
1554 Kumar A, Singh JP, Dwivedi YK, Rana NP (2020) A deep multi-modal neural network
1555 for informative Twitter content classification during emergencies. Annals of Operations
1556 Research DOI: 10.1007/s10479-020-03514-x
1557 Kundu S, Srijith P, Desarkar MS (2018) Classification of short-texts generated dur-
1558 ing disasters: A deep neural network based approach. In: Proceedings of the 2018
1559 IEEE/ACM International Conference on Advances in Social Networks Analysis and
1560 Mining (ASONAM), IEEE, pp 790–793, DOI: 10.1109/ASONAM.2018.8508695
1561 Kusumawardani RP, Hafidz I, Putra SF (2016) BencanaVis visualization and clustering
1562 of disaster readiness using K Means with R shiny a case study for disaster, medical
1563 personnel and health facilities data at province level in Indonesia. In: Proceedings of
1564 the 2016 International Conference on Information & Communication Technology and
1565 Systems (ICTS), IEEE, pp 178–186, DOI: 10.1109/ICTS.2016.7910295
1566 Ladds M, Keating A, Handmer K, Magee L (2017) How much do disasters cost? a comparison
1567 of disaster cost estimates in Australia. International Journal of Disaster Risk Reduction
1568 21:419–429
1569 Lagaros ND, Fragiadakis M (2007) Fragility assessment of steel frames using neural networks.
1570 Earthquake Spectra 23:735–752
1571 Lam NSN, Reams M, Li K, Li C, Mata LP (2016) Measuring community resilience to coastal
1572 hazards along the Northern Gulf of Mexico. Natural Hazards Review 17:04015013
1573 Lattanzi D, Miller G (2017) Review of robotic infrastructure inspection systems. Journal of
1574 Infrastructure Systems 23:04017004
1575 Lattanzi D, Miller GR (2015) 3D scene reconstruction for robotic bridge inspection. Journal
1576 of Infrastructure Systems 21:04014041
1577 Layek AK, Poddar S, Mandal S (2019) Detection of flood images posted on online social
1578 media for disaster response. In: Proceedings of the 2019 Second International Confer-
1579 ence on Advanced Computational and Communication Paradigms (ICACCP), DOI:
1580 10.1109/ICACCP.2019.8882877
1581 Laylavi F, Rajabifard A, Kalantari M (2017) Event relatedness assessment of Twitter mes-
1582 sages for emergency response. Information Processing and Management 53:266–280
1583 Lee J, Eo G, Choi C, Jung J, Kim H (2016) Development of rainfall-flood damage esti-
1584 mation function using nonlinear regression equation. Journal of the Society of Disaster
1585 Information 12:74–88
1586 Lee W, Kim S, Lee YT, Lee HW, Choi M (2017) Deep neural networks for wild fire de-
1587 tection with unmanned aerial vehicle. In: Proceedings of the 2017 IEEE International
1588 Conference on Consumer Electronics (ICCE), IEEE, DOI: 10.1109/ICCE.2017.7889305
1589 Lenjani A, Yeum CM, Dyke SJ, Bilionis I (2019) Automated building image extraction from
1590 360◦ panoramas for postdisaster evaluation. Computer-Aided Civil and Infrastructure
1591 Engineering DOI: 10.1111/mice.12493
1592 Lenz I (2016) Deep learning for robotics. PhD dissertation, Cornell University
1593 Leon F, Atanasiu GM (2006) Data mining methods for GIS analysis of seismic vulnerability.
1594 Proceedings of the First International Conference on Software and Data Technologies
1595 (ICSOFT 2006) 2:153–156
1596 Leśniak A, Isakow Z (2009) Spacetime clustering of seismic events and hazard assessment
1597 in the Zabrze-Bielszowice coal mine, Poland. International Journal of Rock Mechanics
1598 and Mining Sciences 46:918–928
1599 Li H, Parikh D, He Q, Qian B, Li Z, Fang D, Hampapur A (2014) Improving rail net-
1600 work velocity: A machine learning approach to predictive maintenance. Transportation
1601 Research Part C 45:17–26
1602 Li H, Caragea D, Caragea C, Herndon N (2018a) Disaster response aided by tweet classifi-
1603 cation with a domain adaptation approach. Journal of Contingencies and Crisis Man-
 Applications of artificial intelligence for disaster management 41

1604 agement 26:16–27

1605 Li J, Rao H (2010) Twitter as a rapid response news service: an exploration in the con-
1606 text of the 2008 China earthquake. The Electronic Journal of Information Systems in
1607 Developing Countries 42:1–22
1608 Li J, Stephens KK, Zhu Y, Murthy D (2019a) Using social media to call for help in Hurricane
1609 Harvey: Bonding emotion, culture, and community relationships. International Journal
1610 of Disaster Risk Reduction 38:101212
1611 Li Q, Jin Z, Wan C, Zeng DD (2016a) Mining opinion summarizations using convolutional
1612 neural networks in Chinese microblogging systems. Knowledge-Based Systems 107:289–
1613 300
1614 Li S, Teo KL (2018) Post-disaster multi-period road network repair: work scheduling and
1615 relief logistics optimization. Annals of Operations Research 283:1345–1385
1616 Li T, Zhou W, Zeng C, Wang Q, Zhou Q, Wang D, Xue J, Huang Y, Wang W, Zhang
1617 M, Luis S, Chen SC, Rishe N (2016b) DI-DAP: An efficient disaster information deliv-
1618 ery and analysis platform in disaster management. In: Proceedings of the 25th ACM
1619 International on Conference on Information and Knowledge Management, pp 1593–1602
1620 Li T, Xie N, Zeng C, Zhou W, Zheng L, Jiang Y, Yang Y, Ha HY, Xue W, Huang Y, Chen
1621 SC, Navlakha J, Iyengar SS (2017a) Data-driven techniques in disaster information
1622 management. ACM Computing Surveys 50, Article No. 1
1623 Li T, Zeng C, Zhou W, Xue W, Huang Y, Liu Z, Zhou Q, Xia B, Wang Q, Wang W, Zhu
1624 X (2017b) FIU-Miner (a fast, integrated, and user-friendly system for data mining) and
1625 its applications. Knowledge and Information Systems 52:411–443
1626 Li T, Wang Q, Xie Z (2019b) Disaster response knowledge and its social determinants: A
1627 cross-sectional study in Beijing, China. PLoS One 14:e0214367
1628 Li W, Batty M, Goodchild MF (2019c) Real-time GIS for smart cities. International Journal
1629 of Geographical Information Science DOI: 10.1080/13658816.2019.1673397
1630 Li Z, Meier M, Hauksson E, Zhan Z, Andrews J (2018b) Machine learning seismic wave
1631 discrimination: Application to earthquake early warning. Geophysical Research Letters
1632 45:4773–4779
1633 Liang NJ, Shih YT, Shih FY, Wu HM, Wang HJ, Shi SF, Wang BB (2001) Disaster epidemi-
1634 ology and medical response in the Chi-Chi earthquake in Taiwan. Annals of Emergency
1635 Medicine 38(5):549–555, DOI: 10.1067/mem.2001.118999
1636 Liang X (2018) Imagebased postdisaster inspection of reinforced concrete bridge systems us-
1637 ing deep learning with Bayesian optimization. Computer-Aided Civil and Infrastructure
1638 Engineering 34:415–430
1639 Lin GF, Chang MJ, Huang YC, Ho JY (2017a) Assessment of susceptibility to rainfall-
1640 induced landslides using improved self-organizing linear output map, support vector
1641 machine, and logistic regression. Engineering Geology 224:62–74
1642 Lin SY, Chao KM, Lo CC, Godwin N (2013) Distributed dynamic data driven prediction
1643 based on reinforcement learning approach. In: Proceedings of the 28th Annual ACM
1644 Symposium on Applied Computing, ACM, pp 779–784,
1645 Lin TH, Liaw DC (2015) Development of an intelligent disaster information-integrated plat-
1646 form for radiation monitoring. Natural Hazards 76:1711–1725
1647 Lin WT, Chou WC, Lin CY (2008) Earthquake-induced landslide hazard and vegetation
1648 recovery assessment using remotely sensed data and a neural network-based classifier:
1649 a case study in central Taiwan. Natural Hazards 47:331–347
1650 Lin Y, Margolin D, Wen X (2017b) Tracking and analyzing individual distress following
1651 terrorist attacks using social media streams. Risk Analysis 37:1580–1605
1652 Lin YR (2015) Event-related crowd activities on social media. In: Gonalves B, Perra N (eds)
1653 Social Phenomena, Springer, pp 235–250
1654 Lingam G, Rout RR, Somayajulu DVLN (2019) Deep Q-learning and particular swarm
1655 optimization for bot detection in online social networks. In: Proceedings of the 10th
1656 International Conference on Computing, Communication and Networking Technologies
1657 (ICCCNT), DOI: 10.1109/ICCCNT45670.2019.8944493
1658 Liu CH, Chen Z, Tang J, Xu J, Piao C (2018) Energy-efficient UAV control for effective and
1659 fair communication coverage: A deep reinforcement learning approach. IEEE Journal
1660 on Selected Areas in Communications 36:2059–2070
 42 Wenjuan Sun et al.

1661 Liu H, Davidson RA, Apanasovich TV (2008) Spatial generalized linear mixed models of
1662 electric power outages due to hurricanes and ice storms. Reliability Engineering & Sys-
1663 tem Safety 93:897–912
1664 Liu K, Li Z, Yao C, Chen J, Zhang K, Saifullah M (2016) Coupling the k-nearest neighbor
1665 procedure with the Kalman filter for real-time updating of the hydraulic model in flood
1666 forecasting. International Journal of Sediment Research 31:149–158
1667 Liu X, Nourbakhsh A, Li Q, Fang R, Shah S (2015) Real-time rumor debunking on Twit-
1668 ter. In: Proceedings of the 24th ACM International on Conference on Information and
1669 Knowledge Management (CIKM’15), ACM, pp 1867–1870
1670 Liu X, Liu Y, Chen Y (2019a) Reinforcement learning in multiple-UAV networks: Deploy-
1671 ment and movement design. IEEE Transactions on Vehicular Technology 68:8036–8049
1672 Liu Y, Wu L (2016) Geological disaster recognition on optical remote sensing images using
1673 deep learning. Procedia Computer Science 91:566–575
1674 Liu Y, Yang J, Zheng Y, Wu Z, Yao M (2013) Multi-robot coordination in complex envi-
1675 ronment with task and communication constraints. International Journal of Advanced
1676 Robotic Systems 10:229
1677 Liu Z, Zhang Z (2018) Artificial neural network based method for seismic fragility analysis
1678 of steel frames. KSCE Journal of Civil Engineering 22:708–717
1679 Liu Z, Du Y, Yi J, Liang F, Ma T, Pei T (2019b) Quantitative estimates of collective geo-
1680 tagged human activities in response to typhoon Hato using location-aware big data.
1681 International Journal of Digital Earth DOI: 10.1080/17538947.2019.1645894
1682 Lodree EJJ, Davis LB (2016) Empirical analysis of volunteer convergence following the 2011
1683 tornado disaster in Tuscaloosa, Alabama. Natural Hazards 84:1109–1135
1684 Lohumi K, Roy S (2018) Automatic detection of flood severity level from flood videos using
1685 deep learning models. In: Proceedings of the 2018 5th International Conference on Infor-
1686 mation and Communication Technologies for Disaster Management (ICT-DM), IEEE,
1687 DOI: 10.1109/ICT-DM.2018.8636373
1688 Long Z, Wang P, Lin Z, Zhu J (2018) Research on system of marine situation information
1689 analysis and early warning based on artificial intelligence. Advances in Engineering
1690 Research 164:362–368
1691 Lucieer A, de Jong SM, Turner D (2014) Mapping landslide displacements using Structure
1692 from Motion (SfM) and image correlation of multi-temporal UAV photography. Progress
1693 in Physical Geography: Earth and Environment 38:97–116
1694 Luo Y, Liu L, Huang WQ, ad Jie Deng YNY, Yin CH, Ren H, Wang XY (2013) A disaster
1695 response and management competency mapping of community nurses in China. Iranian
1696 Journal of Public Health 42(9):941–949
1697 Ma C, Zhang J, Zhao Y, Habib MF, Savas SS, Mukherjee B (2015a) Traveling repair-
1698 man problem for optical network recovery to restore virtual networks after a disaster.
1699 IEEE/OSA Journal of Optical Communications and Networking 7:B81–B92
1700 Ma X, Tao Z, Wang Y, Yu H, Wang Y (2015b) Long short-term memory neural network for
1701 traffic speed prediction using remote microwave sensor data. Transportation Research
1702 Part C 54:187–197
1703 Mabon L (2016) Charting disaster recovery via Google Street View: A social science per-
1704 spective on challenges raised by the Fukushima nuclear disaster. International Journal
1705 of Disaster Risk Science 7:175–185
1706 MacEachren AM, Jaiswal A, Robinson AC, Pezanowski S, Savelyev A, Mitra P, Zhang X,
1707 Blanford J (2011) Senseplace2: Geotwitter analytics support for situational awareness.
1708 In: Proceedings of the IEEE Conference on Visual Analytics Science and Technology
1709 2011, IEEE, pp 181–190
1710 Maciel-Pearson BG, Marchegiani L, Akcay S, Atapour-Abarghouei A, Garforth J, Breckon
1711 TP (2019) Online deep reinforcement learning for autonomous uav navigation and ex-
1712 ploration of outdoor environments. URL https://arxiv.org/abs/1912.05684
1713 Maharjan L, Ditsworth M, Niraula M, Narvaez CC, Fahimi B (2018) Machine learning based
1714 energy management system for grid disaster mitigation. IET Smart Grid 2:172–182
1715 Mahmoudi SN, Chouinard L (2016) Seismic fragility assessment of highway bridges using
1716 support vector machines. Bulletin of Earthquake Engineering 14:1571–1587
1717 Malawani AD, Nurmandi A, Purnomo EP, Rahman T (2020) Social media in aid of post
1718 disaster management. Transforming Government: People, Process and Policy DOI:
 Applications of artificial intelligence for disaster management 43

1719 10.1108/TG-09-2019-0088
1720 Maliszewski PJ, Larson EK, Perrings C (2012) Environmental determinants of unscheduled
1721 residential outages in the electrical power distribution of Phoenix, Arizona. Reliability
1722 Engineering & System Safety 99:161–171
1723 Mane SS, Mokashi MK (2015) Real-time flash-flood monitoring, alerting and forecasting
1724 system using data mining and wireless sensor network. In: Proceedings of the 2015
1725 International Conference on Communications and Signal Processing (ICCSP), IEEE,
1726 pp 1881–1886, DOI: 10.1109/ICCSP.2015.7322851
1727 Mangalahtu S, Jeon JS, DesRoches R (2018) Critical uncertainty parameters influencing seis-
1728 mic performance of bridges using Lasso regression. Earthquake Engineering & Structural
1729 Dynamics 47:784–801
1730 Mangalathu S, Heo G, Jeon JS (2018) Artificial neural network based multi-dimensional
1731 fragility development of skewed concrete bridge classes. Engineering Structures 162:166–
1732 176
1733 Mangalathu S, Hwang SH, Choi E, Jong-SuJeon (2019) Rapid seismic damage evaluation of
1734 bridge portfolios using machine learning techniques. Engineering Structures 201:109785
1735 Mao H, Alizadeh M, Menache I, Kandula S (2016) Resource management with deep re-
1736 inforcement learning. In: Proceedings of the 15th ACM Workshop on Hot Topics in
1737 Networks, ACM, pp 50–56, DOI: 10.1145/3005745.3005750
1738 Mao H, Thakur G, Sparks K, Sanyal J, Bhaduri B (2019) Mapping near-real-time power
1739 outages from social media. International Journal of Digital Earth 12:1285–1299
1740 Marjanović M, Kovačević M, Bajat B, Voženı́lek V (2011) Landslide susceptibility assess-
1741 ment using SVM machine learning algorithm. Engineering Geology 123:225–234
1742 Mason AJ (2013) Simulation and real-time optimised relocation for improving ambulance
1743 operations. In: Handbook of Healthcare Operations Management, Springer, pp 289–317
1744 Maulik U, Bandyopadhyay S (2002) Performance evaluation of some clustering algorithms
1745 and validity indices. IEEE Transactions on Pattern Analysis and Machine Intelligence
1746 24:1650–1654
1747 McCaslin SE, Jacobs GA, Meyer DL, Johnson-Jimenez E, Metzler TJ, Marmar CR (2005)
1748 How does negative life change following disaster response impact distress among Red
1749 Cross responders? Professional Psychology: Research and Practice 36(3):346–253, DOI:
1750 10.1037/0735-7028.36.3.246
1751 McConnon A (2018) AI helps cities predict natural disasters. https://www.wsj.com/articles/
1752 ai-helps-cities-predict-natural-disasters-1530065100
1753 McCreadie R, Macdonald C, Ounis I (2016) EAIMS: Emergency analysis identification and
1754 management system. In: Proceedings of the 39th International ACM SIGIR conference
1755 on Research and Development in Information Retrieval, ACM, pp 1101–1104, DOI:
1756 10.1145/2911451.2911460
1757 McGovern A, Gagne II DJ, Troutman N, Brown RA, Basara J, Williams JK (2011) Using
1758 spatiotemporal relational random forests to improve our understanding of severe weather
1759 processes. Statistical Analysis and Data Mining 4:407–429
1760 Mehrjoo M, Khaji N, Moharrami H, Bahreininejad A (2008) Damage detection of truss
1761 bridge joints using artificial neural networks. Expert Systems with Applications 35:1122–
1762 1131
1763 Melchiorre C, Matteucci M, Azzoni A, Zanchi A (2008) Artificial neural networks and cluster
1764 analysis in landslide susceptibility zonation. Geomorphology 94:379–400
1765 Memarzadeh M, Pozzi M (2019) Model-free reinforcement learning with model-based safe
1766 exploration: Optimizing adaptive recovery process of infrastructure systems. Structural
1767 Safety 80:46–55
1768 Mendoza M, Poblete B, Castillo C (2010) Twitter under crisis: Can we trust what we RT?
1769 In: The 1st Workshop on Social Media Analytics (SOMA 10), ACM, pp 71–79
1770 Meruane V, Heylen W (2011) Structural damage assessment with antiresonances versus
1771 mode shapes using parallel genetic algorithms. Structural Control Health Monitoring
1772 18:825–839
1773 Merz B, Kreibich H, Lall U (2013) Multi-variate flood damage assessment: a tree-based
1774 data-mining approach. Natural Hazards and Earth System Sciences 13:53–64
1775 Middleton SE, Middleton L, Modafferi S (2014) Real-time crisis mapping of natural disasters
1776 using social media. IEEE Intelligent Systems 29:9–17
 44 Wenjuan Sun et al.

1777 Mihunov VV, Lam NSN, Zou L, Wang Z, Wang K (2020) Use of Twitter in disaster rescue:
1778 lessons learned from Hurricane Harvey. International Journal of Digital Earth DOI:
1779 10.1080/17538947.2020.1729879
1780 Mishra AK, Desai VR (2006) Drought forecasting using feed-forward recursive neural net-
1781 work. Ecological Modelling 198:127–138
1782 Mitropoulou CC, Papadrakakis M (2011) Developing fragility curves based on neural net-
1783 work IDA predictions. Engineering Structures 33:3409–3421
1784 Mitsopoulos I, Mallinis G (2017) A data-driven approach to assess large fire size generation
1785 in Greece. Natural Hazards 88:1591–1607
1786 Mitsova D, Esnard AM, Sapat A, Lai BS (2018) Socioeconomic vulnerability and electric
1787 power restoration timelines in Florida: the case of Hurricane Irma. Natural Hazards
1788 94:689–709
1789 Mitsova D, Escaleras M, Sapat A, Esnard AM, Lamadrid AJ (2019) The effects of in-
1790 frastructure service disruptions and socio-economic vulnerability on hurricane recovery.
1791 Sustainability 11:516
1792 Moon SH, Kim YH, Lee YH, Moon BR (2018) Application of machine learning to an early
1793 warning system for very short-term heavy rainfall. Journal of Hydrology 568:1042–1054
1794 Moradi A, Razmi J, Babazadeh R, Sabbaghnia A (2019) An integrated principal component
1795 analysis and multi-objective mathematical programming approach to agile supply chain
1796 network design under uncertainty. American Institute of Mathematical Sciences 15:855–
1797 879
1798 Mori K, Yamane A, Hayakawa Y, Wada T, Ohtsuki K, Okada H (2013) Development of
1799 emergency rescue evacuation support system (ERESS) in panic-type disasters: Disas-
1800 ter recognition algorithm by support vector machine. IEICE Transactions on Funda-
1801 mentals of Electronics, Communications and Computer Sciences E96-A:649–657, DOI:
1802 10.1587/transfun.E96.A.649
1803 Morito T, Sugiyama O, Kojima R, Nakadai K (2016) Partially shared deep neural network
1804 in sound source separation and identification using a UAV-embedded microphone array.
1805 In: 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS),
1806 IEEE, pp 1299–1304, DOI: 10.1109/IROS.2016.7759215
1807 Moskowitz H, Drnevich P, Ersoy O, Altinkemer K, Chaturvedi A (2011) Using realtime
1808 decision tools to improve distributed decisionmaking capabilities in highmagnitude crisis
1809 situations. Decision Sciences 42:477–493
1810 Mousavi SM, Zhu W, Sheng Y, Beroza GC (2019) CRED: a deep residual network of convo-
1811 lutional and recurrent units for earthquake signal detection. Scientific Reports 9:10267
1812 Moustapha AI, Selmic RR (2007) Wireless sensor network modeling using modified re-
1813 current neural networks: Application to fault detection. In: 2007 IEEE International
1814 Conference on Networking, Sensing and Control, IEEE, pp 313–318, DOI: 10.1109/IC-
1815 NSC.2007.372797
1816 Moya L, Yamazaki F, Liu W, Yamada M (2018) Detection of collapsed buildings from
1817 LiDAR data due to the 2016 Kumamoto earthquake in Japan. Natural Hazards and
1818 Earth System Sciences 18:65–78
1819 Muda Z, Yassin W, Sulaiman MN, Udzir NI (2011) A K-means and Naı̈ve Bayes learning
1820 approach for better intrusion detection. Information Technology Journal 10:648–655
1821 Muhammad K, Ahmad J, Baik SW (2018) Early fire detection using convolutional neural
1822 networks during surveillance for effective disaster management. Neurocomputing 288:30–
1823 42
1824 Murphy R (2014) Introduction. In: Disaster robotics, The MIT Press, pp 1–20
1825 Murphy RR, Stover S (2007) Rescue robots for mudslides: a descriptive study of the 2005
1826 La Conchita mudslide response. Journal of Field Robotics 25:3–16
1827 Murphy RR, Kravitz J, Stover SL, Shoureshi R (2009) Mobile robotics in mine rescue and
1828 recovery. IEEE Robotics & Automation Magazine 16:91–103
1829 Murphy RR, Steimle E, Hall M, Lindemuth M, Trejo D, Hurlebaus S, Medina-Cetina Z,
1830 Slocum D (2011) Robotics-assisted bridge inspection. Journal of Intelligent& Robotic
1831 Systems 64:77–95
1832 Mutlu B, Nefeslioglu HA, Sezer EA, Akcayol MA, Gokceoglu C (2019) An experimental
1833 research on the use of recurrent neural networks in landslide susceptibility mapping.
1834 International Journal of Geo-information 8:578
 Applications of artificial intelligence for disaster management 45

1835 Nabian MA, Meidani H (2018a) Accelerating stochastic assessment of post-earthquake trans-
1836 portation network connectivity via machine-learning-based surrogates. In: The 97th An-
1837 nual Meeting of Transportation Research Board (TRB), TRB
1838 Nabian MA, Meidani H (2018b) Deep learning for accelerated seismic reliability analysis of
1839 transportation networks. Journal of Computer-Aided Civil and Infrastructure Engineer-
1840 ing 33:443–458
1841 Naidu S, Sajinkumar KS, Oommen T, Anuja VJ, Samuel RA (2018) Early warning system
1842 for shallow landslides using rainfall threshold and slope stability analysis. Geoscience
1843 Frontier 9:1871–188
1844 Naito S, Tomozawa H, Mori Y, Nakamura H, Fujiwara H (2018) Damage detection method
1845 for buildings with machine-learning techniques utilizing images of automobile running
1846 surveys aftermath of the 2016 Kumamoto Earthquake. Journal of Disaster Research
1847 13:928–942,
1848 Nateghi R, Guikema SD, Quiring SM (2014) Power outage estimation for tropical cyclones:
1849 Improved accuracy with simpler models. Risk Analysis 34:1069–1078
1850 Neppalli VK, Caragea C, Squicciarini A, Tapia A, Stehle S (2017) Sentiment analysis during
1851 Hurricane Sandy in emergency response. International Journal of Disaster Risk Reduc-
1852 tion 21:213–222
1853 Neppalli VK, Caragea C, Caragea D (2018) Deep neural networks versus Naı̈ve Bayes clas-
1854 sifiers for identifying informative tweets during disasters. In: Boersma K, Tomaszewski
1855 B (eds) Proceedings of the 15th International Conference on Information Systems for
1856 Crisis Response and Management (ISCRAM 2018), ACM, URL http://idl.iscram.org/
1857 files/venkatakishoreneppalli/2018/2141 VenkataKishoreNeppalli etal2018.pdf
1858 Ngiam J, Khosla A, Kim M, Nam J, Nam H, Ng AY (2011) Multimodal deep learning.
1859 In: Proceedings of the 28th International Conference on Machine Learning, ACM, pp
1860 689–696
1861 Nguyen C, Han F, Schlesinger KJ, Gür I, Carlson JM (2016) Collective decision dynamics
1862 in group evacuation: Behavioral experiment and machine learning models. URL https:
1863 //arxiv.org/abs/1606.05647
1864 Nguyen DT, Ofli F, Imran M, Mitra P (2017) Damage assessment from social media imagery
1865 data during disasters. In: Proceedings of the 2017 IEEE/ACM International Conference
1866 on Advances in Social Networks Analysis and Mining 2017, ACM, pp 569–576
1867 Nguyen L, Yang Z, Li J, Pan Z, Cao G, Jin F (2019a) Forecasting people’s needs in hur-
1868 ricane events from social network. IEEE Transactions on Big Data DOI: 10.1109/TB-
1869 DATA.2019.2941887
1870 Nguyen VQ, Anh TN, Yang HJ (2019b) Real-time event detection using recurrent neural
1871 network in social sensors. International Journal of Distributed Sensor Networks 15(6),
1872 DOI: 10.1177/1550147719856492
1873 NHERI (2019) DesignSafe. URL https://www.designsafe-ci.org/
1874 Ning L, Li Y, Zhou M, Song H, Dong H (2019) A deep reinforcement learning approach to
1875 high-speed train timetable rescheduling under disturbances. In: 2019 IEEE Intelligent
1876 Transportation Systems Conference (ITSC), IEEE, pp 3469–3474
1877 NIST (2018) Community resilience planning guide. https://www.nist.gov/topics/
1878 community-resilience/community-resilience-planning-guide
1879 Noda K, Arie H, Suga Y, Ogata T (2014) Multimodal integration learning of robot behavior
1880 using deep neural networks. Robotics and Autonomous Systems 62:721–736
1881 Nolasco-Javier D, Kumar L (2018) Deriving the rainfall threshold for shallow landslide early
1882 warning during tropical cyclones: a case study in northern Philippines. Natural Hazards
1883 90:921–941
1884 Novellino A, Jordan C, Ager G, Bateson L, Fleming C, Confuorto P (2018) Remote sensing
1885 for natural or man-made disasters and environmental changes. In: Geological Disaster
1886 Monitoring Based on Sensor Networks, Springer, pp 23–31
1887 Noymanee J, Nikitin NO, Kalyuzhnaya AV (2017) Urban pluvial flood forecasting using open
1888 data with machine learning techniques in Pattani Basin. Procedia Computer Science
1889 119:288–297
1890 Ofli F, Meier P, Castillo C, Tuisa D, Briant J, Millet P, Reinhard F, Parkan M, Joost S
1891 (2016) Combining human computing and machine learning to make sense of Big (aerial)
1892 Data for disaster response. Big Data 4:47–59
 46 Wenjuan Sun et al.

1893 Oh J, Hwang JE, Smith SF (2006) Agent technologies for post-disaster urban planning. In:
1894 First International AAMAS Workshop on Agent Technology for Disaster Management,
1895 pp 24–31
1896 Oktarina R, Bahagia SN, Diawati L, Pribadi KS (2019) Artificial neural network for pre-
1897 dicting earthquake casualties and damages in Indonesia. In: The 3rd International Con-
1898 ference on Eco Engineering, IOPscience, p 012156
1899 Orabi W, Senouci AB, El-Rayes K, Al-Derham H (2010) Optimizing resource utilization dur-
1900 ing the recovery of civil infrastructure systems. Journal of Management in Engineering
1901 26:237–246
1902 Otoum S, Kantarci B, Mouftah H (2019) Empowering reinforcement learning on big sensed
1903 data for intrusion detection. In: Proceedings of the 2019 IEEE International Conference
1904 on Communications (ICC2019), IEEE, DOI: 10.1109/ICC.2019.8761575
1905 Özdamar L, Demir O (2012) A hierarchical clustering and routing procedure for large scale
1906 disaster relief logistics planning. Transportation Research Part E: Logistics and Trans-
1907 portation Review 48:591–602
1908 Ozdemir A, Altural T (2013) A comparative study of frequency ratio, weights of evidence
1909 and logistic regression methods for landslide susceptibility mapping: Sultan Mountains,
1910 SW Turkey. Journal of Asian Earth Sciences 64:180–197
1911 Padil KH, Bakhary N, Hao H (2017) The use of a nonprobabilistic artificial neural network
1912 to consider uncertainties in vibration-based-damage detection. Mechanical Systems and
1913 Signal Processing 83:194–209
1914 Padmawar PM, Shinde AS, Sayyed TZ, Shinde SK, Moholkar K (2019) Disaster predic-
1915 tion system using convolution neural network. In: The 2019 International Conference on
1916 Communication and Electronics Systems (ICCES), IEEE, pp 808–812, DOI: 10.1109/IC-
1917 CES45898.2019.9002400
1918 Park YS, Kim J, Kim A (2019) Radar localization and mapping for indoor disaster environ-
1919 ments via multi-modal registration to prior LiDAR map. In: 2019 IEEE/RSJ Interna-
1920 tional Conference on Intelligent Robots and Systems (IROS), IEEE, pp 407–419, DOI:
1921 10.1109/IROS40897.2019.8967633
1922 Parra J, Fuentes O, Anthony EY, Kreinovich V (2016) Use of machine learning to analyze
1923 and – hopefully – predict volcano activity. Tech. Rep. UTEP-CS-16-80a, University of
1924 Texas at El Paso, http://digitalcommons.utep.edu/cs techrep/1053
1925 Pau J, Baker J, Houston N (2017) Artificial intelligence in Asia: Preparedness and resilience.
1926 URL https://www.asiabusinesscouncil.org/docs/AI briefing.pdf
1927 Pechenkin A, Demidov R (2018) Application of deep neural networks for security analysis
1928 of digital infrastructure components. SHS Web of Conferences 44:00068
1929 Peduzzi P, Dao H, Herold C, Mouton F (2009) Assessing global exposure and vulnerability
1930 towards natural hazards: the Disaster Risk Index. Natural Hazards and Earth System
1931 Sciences 9:1149–1159
1932 Peiris RH, Hallé C, Budman H, Moresoli C, Peldszus S, Huck PM, Legge RL (2010) Identi-
1933 fying fouling events in a membrane-based drinking water treatment process using prin-
1934 cipal component analysis of fluorescence excitation-emission matrices. Water Research
1935 44:185–194
1936 Peng Y, Li SW, Hu ZZ (2019) A self-learning dynamic path planning method for evacuation
1937 in large public buildings based on neural networks. Neurocomputing 365:71–85
1938 Perol T, Gharbi M, Denolle M (2018) Convolutional neural network for earthquake detection
1939 and location. Science Advances 4:e1700578
1940 Pessin G, Osorio F, Wolf DF, Dias MA (2009) Genetic algorithm applied to robotic squad
1941 coordination. In: 2009 Electronics, Robotics and Automotive Mechanics Conference
1942 (CERMA), IEEE, pp 169–174, DOI: 10.1109/CERMA.2009.17
1943 Pezanowski S, MacEachren AM, Savelyev A, Robinson AC (2018) SensePlace3: a geovi-
1944 sual framework to analyze placetimeattribute information in social media. Journal of
1945 Cartography and Geographic Information Science 45:420–437
1946 Pham BT, Bui DT, Prakash I, Dholakia MB (2017) Hybrid integration of multilayer per-
1947 ceptron neural networks and machine learning ensembles for landslide susceptibility
1948 assessment at Himalayan area (India) using GIS. CATENA 149:52–63
1949 Pham TTH, Apparicio P, Gomez C, Weber C, Mathon D (2014) Towards a rapid automatic
1950 detection of building damage using remote sensing for disaster management: The 2010
 Applications of artificial intelligence for disaster management 47

1951 Haiti Earthquake. Disaster Prevention and Management 23:53–66

1952 Pilkington SF, Mahmoud HN (2016) Using artificial neural networks to forecast economic
1953 impact of multi-hazard hurricane-based events. Sustainable and Resilient Infrastructure
1954 1:63–83
1955 Pillai AS, Chandraprasad GS, Khwaja AS, Anpalagan A (2019) A service oriented IoT
1956 architecture for disaster preparedness and forecasting system. Internet of Things DOI:
1957 10.1016/j.iot.2019.100076
1958 Pogrebnykov N, Maldonado E (2017) Identifying emergency stages in Facebook posts of
1959 police departments with convolutional and recurrent neural networks and support vector
1960 machines. In: Proceedings of the 2017 IEEE International Conference on Big Data,
1961 IEEE, pp 4343–4352, DOI:10.1109/BigData.2017.8258464
1962 Pourrahmani E, Delavar MR, Mostafavi MA (2015) Optimization of an evacuation plan with
1963 uncertain demands using fuzzy credibility theory and genetic algorithm. International
1964 Journal of Disaster Risk Reduction 14:357–372
1965 Pouyanfar S, Sadiq S, Tian H, Tao Y, Reyes MP, ling Shyu M, Chen SC, Iyengar SS (2018)
1966 A survey on deep learning: algorithms, techniques, and applications. ACM Computing
1967 Surveys 51, article No. 92
1968 Pradhan B (2009) Flood susceptible mapping and risk area delineation using logistic regres-
1969 sion, GIS and remote sensing. Journal of Spatial Hydrology 9:1–18
1970 Pual2012 (2012) Location-allocation planning of stockpiles for effective disaster mitigation.
1971 Annals of Operations Research 196:469–490
1972 Pugh J, Martinoli A (2007) Inspiring and modeling multi-robot search with particle
1973 swarm optimization. In: 2007 IEEE Swarm Intelligence Symposium, IEEE, DOI:
1974 10.1109/SIS.2007.367956
1975 Pyayt AL, Mokhov II, Lang B, Krzhizhanovskaya VV, Meijer RJ (2011) Machine learning
1976 methods for environmental monitoring and flood protection. International Journal of
1977 Computer, Electrical, Automation, Control and Information Engineering 5:549–554
1978 Qiang Y, Huang Q, Xu J (2020) Observing community resilience from space: using night-
1979 time lights to model economic disturbance and recovery pattern in natural disaster.
1980 Sustainable Cities and Society 57:102115
1981 Qiao L, Luo J (2012) Research on Q-learning algorithm with sharing experi-
1982 ence in learning process. Computer Science 39, http://en.cnki.com.cn/Article en/
1983 CJFDTotal-JSJA201205053.htm
1984 Qiu M, Ming Z, Wang J, Yang LT, Xiang Y (2014) Enabling cloud computing in emergency
1985 management systems. IEEE Cloud Computing 1:60–67
1986 Rafiei MH, Adeli H (2017) NEEWS: A novel earthquake early warning model using neural
1987 dynamic classication and neural dynamic optimization. Soil Dynamics and Earthquake
1988 Engineering 100:417–427
1989 Rahman R, Hasan S (2018) Short-term traffic speed prediction for freeways during
1990 hurricane evacuation: A deep learning approach. In: The 21st International Con-
1991 ference on Intelligent Transportation Systems (ITSC), IEEE, pp 1291–1296, DOI:
1992 10.1109/ITSC.2018.8569443
1993 Rahnemoonfar M, Murphy R, Miquel MV, Dobbs D, Adams A (2018) Flooded area detection
1994 from UAV images based on densely connected recurrent neural networks. In: Proceedings
1995 of the 2018 IEEE International Geoscience and Remotely Sensing Symposium, IEEE,
1996 pp 1788–1791, DOI: 10.1109/IGARSS.2018.8517946
1997 Rakgase M, Norris D (2014) Factors that influence choice of drought coping strategies in
1998 Limpopo Province, South Africa. Journal of Human Ecology 47:111–116
1999 Ramchurn SD, Huynh TD, Ikuno Y, Flann J, Wu F, Moreau L, Jennings NR, Fischer J, Jiang
2000 W, Rodden T, Simpson E, Reece S, Roberts SJ (2015) HAC-ER: A disaster response
2001 system based on human-agent collectives. In: Proceedings of the 2015 International
2002 Conference on Autonomous Agents and Multiagent Systems (AAMAS 2015), pp 533–
2003 541
2004 Ramchurn SD, Huynh TD, Wu F, Ikuno Y, Flann J, Moreau L, Fischer JE, Jiang W, Rodden
2005 T, Simpson E, Reece S, Roberts S, Jennings NR (2016) A disaster response system based
2006 on human-agent collectives. Journal of Artificial Intelligence Research 57:661–708, DOI:
2007 10.1613/jair.5098
 48 Wenjuan Sun et al.

2008 Ranzato M, Mnih V, Susskind JM, Hinton GE (2013) Modeling natural images using gated
2009 MRFs. IEEE Transactions on Pattern Analysis and Machine Intelligence 35:2206–2222
2010 Rasouli MR (2018) Intelligent process-aware information systems to support agility in dis-
2011 aster relief operations: a survey of emerging approaches. International Journal of Pro-
2012 duction Research 57:1857–1872
2013 Rauter M, Winkler D (2018) Predicting natural hazards with neuronal networks. https:
2014 //arxiv.org/pdf/1802.07257.pdf
2015 Rawat P, Haddad M, Altman E (2015) Towards efficient disaster management: 5G and
2016 device to device communication. In: Proceedings of the 2nd International Conference
2017 on Information and Communication Technologies for Disaster Management (ICT-DM),
2018 IEEE, DOI: 10.1109/ICT-DM.2015.7402056
2019 Raymond R, Morimura T, Osogami T, Hirosue N (2012) Map matching with hidden Markov
2020 model on sampled road network. In: Proceedings of the 21st International Conference
2021 on Pattern Recognition (ICPR2012), IEEE, pp 2242–2245
2022 Reed D (2008) Electric utility distribution analysis for extreme winds. Journal of Wind
2023 Engineering & Industrial Aerodynamics 96:123–140
2024 Reilly J, Dashti S, Ervasti M, Bray JD, Glaser SD, Bayen AM (2013) Mobile phones as
2025 seismologic sensors: Automating data extraction for the iShake system. IEEE Robotics
2026 and Automation Society 10:242 – 251
2027 Ren M, Wang B, Fu QLG (2010) Classified real-time flood forecasting by coupling fuzzy
2028 clustering and neural network. International Journal of Sediment Research 25:134–148
2029 Renwick N (2017) China’s approach to disaster risk reduction: Human security challenges
2030 in a time of climate change. Journal of Asian Security and International Affairs 4:26–49
2031 Resch B, Usländer F, Havas C (2018) Combining machine-learning topic models and spa-
2032 tiotemporal analysis of social media data for disaster footprint and damage assessment.
2033 Journal of Artificial Intelligence Research 45:362–376
2034 Reynard D, Shirgaokar M (2019) Harnessing the power of machine learning: Can Twit-
2035 ter data be useful in guiding resource allocation decisions during a natural disaster?
2036 Transportation Research Part D: Transport and Environment 77:449–463
2037 Rhee J, Im J (2017) Meteorological drought forecasting for ungauged areas based on machine
2038 learning: using long-range climate forecast and remote sensing data. Agricultural and
2039 Forest Meteorology 237-238:105–122
2040 Riad JK, Norris FH, Ruback RB (2006) Predicting evacuation in two major disasters: Risk
2041 perception, social influence, and access to resources. Journal of Applied Social Psychol-
2042 ogy 29:918–934
2043 Rizk Y, Jomaa H, Award M, Castillo C (2019) A computationally efficient multi-modal
2044 classification approach of disaster-related Twitter images. In: The 34th ACM/SIGAPP
2045 Symposium on Applied Computing (SAC 19), ACM, DOI: 10.1145/3297280.3297481
2046 Robertson BW, Johnson M, Murthy D, Smith WR, Stephens KK (2019) Using a combination
2047 of human insights and deep learning for real-time disaster communication. Progress in
2048 Disaster Science 2:100030
2049 Robinson S, Murphy H, Bies A (2014) Structured to partner: School district collaboration
2050 with nonprofit organizations in disaster response. Risk, Hazards & Crisis 5(1):77–95,
2051 DOI: 10.1002/rhc3.12047
2052 Rodriguez-Ramos A, Sampedro C, Bavle H, de la Puente P, Campoy P (2019) A deep
2053 reinforcement learning strategy for UAV autonomous landing on a moving platform.
2054 Journal of Intelligent & Robotics Systems 93:351–366
2055 Romlay MRM, Rashid MM, Toha SF (2016) Development of particle swarm optimization
2056 based rainfall-runoff prediction model for Pahang River Pekan. In: Proceedings of the
2057 2016 International Conference on Computer and Communication Engineering (ICCCE),
2058 IEEE, pp 306–310, DOI: 10.1109/ICCCE.2016.72
2059 Rosellini AJ, Dussailant F, Zubizarreta JR, Kessler RC, Rose S (2018) Predicting post-
2060 traumatic stress disorder following a natural disaster. Journal of Psychiatric Research
2061 96:15–22
2062 Rosser JF, Leibovici DG, Jackson MJ (2017) Rapid flood inundation mapping using social
2063 media, remote sensing and topographic data. Natural Hazards 87:103–120
2064 Rossi F (2019) Building trust in artificial intelligence. Journal of International Affairs
2065 72(1):127–134
 Applications of artificial intelligence for disaster management 49

2066 Ruan JH, Wang XP, Chan FTS, Shi Y (2016) Optimizing the intermodal transportation
2067 of emergency medical supplies using balanced fuzzy clustering. International Journal of
2068 Production Research 54:4368–4386
2069 Rudin C, Waltz D, Anderson RN, Boulanger A, Salleb-Aouissi A, Chow M, Dutta H, Gross
2070 P, Huang B, Ierome S, Isaac D, Kressner A, Passonneau RJ, Radeva A, Wu L (2012)
2071 Machine learning for the New York City power grid. IEEE Transactions on Pattern
2072 Analysis and Machine Intelligence 34:328–345
2073 Rudner TGJ, Ruβwurm M, Fil J, Pelich R, Bischke B, Kopačková V, Biliński P (2019)
2074 Multi3Net: Segmenting flooded buildings via fusion of multiresolution, multisensor, and
2075 multitemporal satellite imagery. In: The Thirty-Third AAAI Conference on Artificial
2076 Intelligence (AAAI-19), AAAI, pp 702–709
2077 Russell SJ, Norvig P (2016) Learning from examples. In: Artificial Intelligence: A Modern
2078 Approach, Third Edition, Pearson, chap 18, pp 693–767
2079 Ruz GA, Henrı́quez PA, no AM (2020) Sentiment analysis of Twitter data during criti-
2080 cal events through Bayesian networks classifiers. Future Generation Computer Systems
2081 106:92–104
2082 Saad OM, Shalaby A, Sayed MS (2014) Automatic arrival time detection for earthquakes
2083 based on fuzzy possibilistic C-means clustering algorithm. In: Proceedings of the 2017
2084 8th International Conference on Recent Advances in Space Technologies (RAST), IEEE,
2085 pp 1749–1758
2086 Sachdeva S, Bhatia T, Verma AK (2018) GIS-based evolutionary optimized gradient boosted
2087 decision trees for forest fire susceptibility mapping. Natural Hazards 92:1399–1418
2088 Sadhu V, Salles-Loustau G, Pompili D, Zonouz S, Sritapan V (2017) Argus: Smartphone-
2089 enabled human cooperation for disaster situational awareness via MARL. In: Proceed-
2090 ings of the 2017 IEEE International Conference on Pervasive Computing and Commu-
2091 nications Demonstrations, IEEE, DOI: 10.1109/PERCOMW.2017.7917529
2092 Sadiq FI, Selamat A, Ibrahim R (2015) Human activity recognition prediction for crowd dis-
2093 aster mitigation. In: Asian Conference on Intelligent Information and Database Systems
2094 (ACIID), Springer, pp 200–210
2095 Sadiq FI, Selamat A, Ibrahim R (2018) Performance valuation of classifiers on activity
2096 recognition for disasters mitigation using smartphone sensing. Jurnal Teknologi 77:11–
2097 19
2098 Saito H, Nakayama D, Matsuyama H (2009) Comparison of landslide susceptibility based on
2099 a decision-tree model and actual landslide occurrence: the Akaishi Mountains, Japan.
2100 Geomorphology 109:108–121
2101 Sakaki T, Okazaki M, Matsuo Y (2012) Tweet analysis for real-time event detection and
2102 earthquake reporting system development. IEEE Transactions on Knowledge and Data
2103 Engineering 25:919–931
2104 Salmane H, Khoudour L, Ruicheck Y (2015) A video-analysis-based railway-road safety sys-
2105 tem for detecting hazard situations at level crossings. IEEE Transactions on Intelligent
2106 Transportation Systems 16:596–609
2107 Samir M, Chraiti M, Assi C, Ghrayeb A (2019) Joint optimization of UAV trajectory and
2108 radio resource allocation for drive-thru vehicular networks. In: Proceedings of the 2019
2109 IEEE Wireless Communications and Networking Conference (WCNC), IEEE, DOI:
2110 10.1109/WCNC.2019.8885893
2111 Sánchez-Garcı́a J (2019) A distributed PSO-based exploration algorithm for a UAV network
2112 assisting a disaster scenario. Future Generation Computer Systems 90:129–148
2113 Sankaranarayanan S, Prabhakar M, Satish S, Jain P, Ramprasad A, Krishnan A (2019)
2114 Flood prediction based on weather parameters using deep learning. Journal of Water
2115 and Climate Change DOI: 10.2166/wcc.2019.321
2116 Sarabakha A, Kayacan E (2016) Y6 tricopter autonomous evacuation in an indoor environ-
2117 ment using Q-learning algorithm. In: The 2016 IEEE 55th Conference on Decision and
2118 Control (CDC), IEEE, pp 5992–5997, DOI: 10.1109/CDC.2016.7799189
2119 Sarkale Y, Nozhati S, Chong EKP, Ellingwood BR, Mahmoud H (2018) Solving Markov
2120 decision processes for network-level post-hazard recovery via simulation optimization
2121 and rollout. In: The IEEE 14th International Conference on Automation Science and
2122 Engineering (CASE), IEEE, pp 906–912
 50 Wenjuan Sun et al.

2123 Saxena A, Wong LLS, Ng AY (2008) Learning grasp strategies with partial shape informa-
2124 tion. In: Proceedings of the 23rd national conference on Artificial intelligence - Volume
2125 3, pp 1491–1494
2126 Schempf H, Chemel B, Everett N (1995) Neptune: Above-ground storage tank inspection
2127 robot system. IEEE Robotics and Automation Society Magazine 2:6–15
2128 Schneider FE, Wildermuth D (2017) Using robots for firefighters and first responders:
2129 scenario specificaton and exermplary system description. In: Proceedings of the 18th
2130 Carpathian Control Conference, pp 216–221
2131 Schwartz J (2018) How can AI help to prepare for floods in a
2132 climate-changed world? URL https://www.scientificamerican.com/article/
2133 former-fema-chief-uses-ai-to-prepare-for-hurricanes-and-rising-seas/
2134 Seydi S, Rastiveis H (2019) A deep learning framework for roads network damage assess-
2135 ment using post-earthquake LiDAR data. International Society for Photogrammetry
2136 and Remote Sensing XLII-4/W18:955–961
2137 Sharma H, Anderson PA, Granmo OC, Goodwin M (2020) Deep Q-learning with Q-matrix
2138 transfer learning for novel fire evacuation. IEEE Transactions on Systems, Man, and
2139 Cybernetics: Systems DOI: 10.1109/TSMC.2020.2967936
2140 Sharma S, Ogunlana K (2015) Using genetic algorithm & neural network for modeling
2141 learning behavior in a multi-agent system during emergency evacuation. International
2142 Journal for Computers & Their Applications 22:172–182
2143 Shelton T, Poorthuis A, Graham M, Zook M (2014) Mapping the data shadows of Hurricane
2144 Sandy: Uncovering the sociospatial dimensions of big data. Geoforum 52:167–179
2145 Sheu JB (2007) An emergency logistics distribution approach for quick response to urgent
2146 relief demand in disasters. Transportation Research Part E: Logistics and Transportation
2147 Review 43:687–709
2148 Sheu JB (2010) Dynamic relief-demand management for emergency logistics operations un-
2149 der large-scale disasters. Transportation Research Part E: Logistics and Transportation
2150 Review 46:1–17
2151 Sheykhmousa M, Kerle N, Kuffer M, Ghaffarian S (2019) Post-disaster recovery assessment
2152 with machine learning-derived land cover and land use information. Remote Sensing
2153 11:1174
2154 Shi H, Gao Q, Qi Y, Liu J, Hu Y (2010) Wind erosion hazard assessment of the mongolian
2155 plateau using FCM and GIS techniques. Environmental Earth Sciences 61:689–697
2156 Shi H, Kim M, Lee SC, Pyo S, Esfahani I, Yoo C (2015) Localized indoor air quality
2157 monitoring for indoor pollutants healthy risk assessment using sub-principal component
2158 analysis driven model and engineering big data. Korean Journal of Chemical Engineering
2159 32:1960–1969
2160 Shibuya Y, Tanaka H (2019) Using social media to detect socio-economic disaster recovery.
2161 IEEE Intelligent Systems 34:29–37
2162 Shirzadi A, Bui DT, Pham BT, Solaimani K, Chapi K, Kavian A, Shahabi H, Revhaug
2163 I (2017) Shallow landslide susceptibility assessment using a novel hybrid intelligence
2164 approach. Environmental Earth Sciences 76:60
2165 Shirzaei M, Walter TR (2010) Timedependent volcano source monitoring using interfero-
2166 metric synthetic aperture radar time series: A combined genetic algorithm and Kalman
2167 filter approach. Journal of Geophysical Research 115, article No.: B10421
2168 da Silva S, Júnior MD, Junior VL, Brennan MJ (2008) Structural damage detection by
2169 fuzzy clustering. Mechanical Systems and Signal Processing 22:1636–1649
2170 Silver D, Lever G, Heess N, Degris T, Wierstra D, Riedmiller M (2014) Deterministic policy
2171 gradient algorithms. In: Proceedings of the 31st International Conference on Machine
2172 Learning (ICML-14), ACM, pp 387–395
2173 Simmons KM, Sutter D (2008) Tornado warning, lead times and tornado casualties: an
2174 empirical investigation. Weather and Forecasting 23:246–258
2175 Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale visual
2176 recognition. URL https://arxiv.org/abs/1409.1556
2177 Singh N, Roy N, Gangopadhyay A (2019) Analyzing the emotions of crowd for improving
2178 the emergency response services. Pervasive and Mobile Computing 58:101018
2179 Smith AB, Katz RW (2013) US billion-dollar weather and climate disasters: data sources,
2180 trends, accuracy and biases. Natural Hazards 67:387–410
 Applications of artificial intelligence for disaster management 51

2181 Socher R, Huval B, Bhat B, Manning CD, Ng AY (2008) Convolutional-recursive deep

2182 learning for 3D object classification. In: Proceedings of the 25th International Conference
2183 on Neural Information Processing Systems - Volume 1, pp 656–664
2184 Socher R, Lin CC, Manning C, Ng AY (2011) Parsing natural scenes and natural language
2185 with recursive neural networks. In: Proceedings of the International Conference on Ma-
2186 chine Learning, Omnipress, pp 129–136
2187 Song X, Zhang Q, Sekimoto Y, Shibasaki R (2014) Prediction of human emergency behav-
2188 ior and their mobility following large-scale disaster. In: Proceedings of the 20th ACM
2189 SIGKDD International Conference on Knowledge Discovery and Data Mining, ACM,
2190 pp 5–14
2191 Song X, Zhang Q, Sekimoto Y, Shibasaki R, Yuan NJ, Xie X (2015) A simulator of human
2192 emergency mobility following disasters: Knowledge transfer from big disaster data. In:
2193 Twenty-Ninth AAAI Conference on Artificial Intelligence, AAAI, pp 730–736
2194 Song X, Zhang Q, Sekimoto Y, Shibasaki R, Yuan NJ, Xie X (2016) Prediction and simu-
2195 lation of human mobility following natural disasters. ACM Transactions on Intelligent
2196 Systems and Technology 8, Article No. 29
2197 Song X, Shibasaki R, Yuan NJ, Xie X, Li T, Adachi R (2017) DeepMob: Learning deep
2198 knowledge of human emergency behavior and mobility from big and heterogeneous data.
2199 ACM Transactions on Intelligent Systems and Technology 35, Article No. 41
2200 Sriram LMK, Ulak MB, Ozguven EE, Arghandeh R (2019) Multi-network vulnerabil-
2201 ity causal model for infrastructure co-resilience. IEEE Access 7:35344–35358, DOI:
2202 10.1109/ACCESS.2019.2904457
2203 Srivastava M, Abdelzaher T, Szymanski B (2012) Human-centric sensing. Philosophical
2204 Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
2205 370:176–197
2206 Stefanov WL, Evans CA (2014) The international space station: A unique platform for
2207 remote sensing of natural disasters. Tech. Rep. 20150003831, NASA, https://ntrs.nasa.
2208 gov/search.jsp?R=20150003831
2209 Steimle ET, Murphy RR, Lindemuth M, Hall ML (2009) Unmanned marine vehi-
2210 cle use at Hurricanes Wilma and Ike. In: OCEANS 2009, IEEE, pp 1–6, DOI:
2211 10.23919/OCEANS.2009.5422201
2212 Stojadinovic Z, Kovacevic M, Marinkovic D, Stojadinovic B (2017) Data-driven hous-
2213 ing damage and repair cost prediction framework based on the 2010 Kraljevo earth-
2214 quake data. In: Proceedings of the 16th World Conference on Earthquake Engineering
2215 (16WCEE), p 4987,
2216 Strauss A, Hoffman S, Wendner R, Bergmeiter K (2009) Structural assessment and reliability
2217 analysis for existing engineering structures, applications for real structures. Structure
2218 and Infrastructure Engineering 5:277–286
2219 Su S, Pi J, Wan C, Li H, Xiao R, Li B (2015) Categorizing social vulnerability patterns in
2220 Chinese coastal cities. Ocean & Coastal Mangement 116:1–8
2221 Su Z, Jiang J, Liang C, Zhang G (2011) Path selection in disaster response management
2222 based on Q-learning. International Journal of Automation and Computing 8:100–106
2223 Suganya R, Jayashree LS (2018) Fuzzy rough set inspired rate adaptation and resource
2224 allocation using Hidden Markov Model (FRSIRA-HMM) in mobile ad hoc networks.
2225 Cluster Computing 22:9875–9888
2226 Sun H, Tian X, Li Z, Chen E, Wang W (2017) Remotely sensed monitoring for-
2227 est changes-a case study in the Jinhe town of Inner Mongolia. In: 2017 IEEE
2228 International Geoscience and Remote Sensing Symposium (IGARSS), IEEE, DOI:
2229 10.1109/IGARSS.2017.8127747
2230 Sun H, Burton H, Wallace J (2019) Reconstructing seismic response demands across multiple
2231 tall buildings using kernel–based machine learning methods. Structural Control Health
2232 Monitoring 26:e2359
2233 Sun W, Bocchini P, Davison BD (2020a) Model for estimating the impact of
2234 interdependencies on system recovery. Journal of Infrastructure Systems DOI:
2235 10.1061/(ASCE)IS.1943-555X.0000569
2236 Sun W, Bocchini P, Davison BD (2020b) Resilience metrics and measurement methods for
2237 transportation infrastructure: the state of the art. Sustainable and Resilient Infrastruc-
2238 ture 5(3):168–199, DOI: 10.1080/23789689.2018.1448663
 52 Wenjuan Sun et al.

2239 Sun W, Bocchini P, Davison BD (2021) Quantitative models for interdependent functional-
2240 ity and recovery of critical infrastructure systems. In: Objective Resilience: Manual of
2241 Practice, ASCE, under review
2242 Sun Y, Li S (2016) Real-time collaborative GIS: a technological review. ISPRS Journal of
2243 Photogrammetry and Remote Sensing 115:143–152
2244 Sun Y, Tan W (2019) A trust-aware task allocation method using deep Q–learning for
2245 uncertain mobile crowdsourcing. Human-centric Computing and Information Sciences
2246 9, article No.: 25
2247 Suriya M, Sumithra MG (2019) Enhancing cooperative spectrum sensing in flying cell towers
2248 for disaster management using convolutional neural networks. In: Proceedings of the EAI
2249 International Conference on Big Data Innovation for Sustainable Cognitive Computing,
2250 pp 181–190
2251 Sutton RS, Barto AG (2018) 6. temporal-difference learning. In: Sutton RS, Barto AG (eds)
2252 Reinforcement learning: an introduction, second edition, The MIT Press, pp 119–140
2253 Takahashi B, Jr ECT, Carmichael C (2015) Communicating on Twitter during a disaster:
2254 An analysis of tweets during Typhoon Haiyan in the Philippines. Computers in Human
2255 Behavior 50:392–398
2256 Takeda T, Mori Y, Kubota N, Arai Y (2014) A route planning for disaster waste disposal
2257 based on robot technology. In: 2014 IEEE Symposium on Robotic Intelligence in Infor-
2258 mationally Structured Space (RiiSS), IEEE, DOI: 10.1109/RIISS.2014.7009173
2259 Tan D, Qu W, Tu J (2010) The damage detection based on the fuzzy clustering and support
2260 vector machine. In: The 2010 International Conference on Intelligent System Design and
2261 Engineering Application, pp 598–601, DOI: 10.1109/ISDEA.2010.404
2262 Tanaka G, Yamane T, Héroux JB, Nakane R, Kanazawa N, Takeda S, Numata H, Nakano
2263 D, Hirose A (2019) Recent advances in physical reservoir computing: A review. Neural
2264 Networks 115:100–123
2265 Tapia AH, Bajpai K, Jansen BJ, Yen J (2011) Seeking the trustworthy tweet: Can mi-
2266 croblogged data fit the information needs of disaster response and humanitarian relief
2267 organizations. In: The 8th International Conference on Information Systems for Crisis
2268 Response and Management: From Early-Warning Systems to Preparedness and Training
2269 (ISCRAM 2011), pp 1–10
2270 Tapia C, Padgett JE (2015) Multi-objective optimisation of bridge retrofit and post-event re-
2271 pair selection to enhance sustainability. Structure and Infrastructure Engineering 12:93–
2272 107
2273 Tatem AJ, Qiu Y, Smith DL, Sabot O, Ali AS, Moonen B (2009) The use of mobile phone
2274 data for the estimation of the travel patterns and imported Plasmodium falciparum
2275 rates among Zanzibar residents. Malaria Journal 8:287
2276 Tatsubori M, Watanabe H, Shibayama A, Sato S, Imamura F (2012) Social web in disaster
2277 archives. In: The Proceedings of the 21st International Conference Companion on World
2278 Wide Web - WWW’12 Companion, ACM, pp 715–716, DOI: 10.1145/2187980.2188190
2279 Terranova OG, Gariano SL, Iaquinta P, Iovine GGR (2015) GA SAKe: forecasting land-
2280 slide activations by a genetic-algorithms-based hydrological model. Geoscientific Model
2281 Development 8:1955–1978
2282 The PRAISys Team (2018) Probabilistic Resilience Assessment of Interdependent Systems
2283 (PRAISys). http://praisys.org
2284 The Rockefeller Foundation (2019) 100 Resilient Cities (100RC). http://www.
2285 100resilientcities.org
2286 Tian H, Chen SC (2017a) MCA-NN: Multiple correspondence analysis based neural net-
2287 work for disaster information detection. In: The 3rd IEEE International Conference on
2288 Multimedia Big Data, IEEE, DOI: 10.1109/BigMM.2017.30
2289 Tian H, Chen SC (2017b) A video-aided semantic analytics system for disaster information
2290 integration. In: The 3rd IEEE International Conference on Multimedia Big Data, IEEE,
2291 pp 242–243, DOI: 10.1109/BigMM.2017.31
2292 Tian H, Pouyanfar S, Chen J, Chen SC, Iyengar SS (2018) Automatic convolutional neural
2293 network selection for image classification using genetic algorithms. In: 2018 IEEE In-
2294 ternational Conference on Information Reuse and Integration (IRI), pp 444–451, DOI:
2295 10.1109/IRI.2018.00071
 Applications of artificial intelligence for disaster management 53

2296 Tinoco J, Correia AG, Cortez P, Toll DG (2018) Data-driven model for stability condition
2297 prediction of soil embankments based on visual data features. Journal of Computing in
2298 Civil Engineering 32:04018027
2299 Tinoco J, Correia AG, Cortez P, Toll D (2019) Combining artificial neural networks and
2300 genetic algorithms for rock cuttings slopes stability condition identification. In: Interna-
2301 tional Conference on Inforatmion technology in Geo-Engineering, Springer, pp 196–209
2302 Tomaszewski B, Blanford J, Ross K, Pezanowski S, MacEachren AM (2011) Supporting
2303 geographically-aware web document foraging and sensemaking. Computers, Environ-
2304 ment and Urban Systems 35:192–207
2305 Tomin N, Kurbatsky V, Rehtanz C (2013) An intelligent security alert system for power
2306 system pre-emergency control. In: The 13th International Conference on Environment
2307 and Electrical Engineering (EEEIC), IEEE, DOI: 10.1109/EEEIC-2.2013.6737884
2308 Toreyin BU, Cetin AE (2009) Wildfire detection using LMS based active learning. In: Pro-
2309 ceedings of the 2009 IEEE International Conference on Acoustics, Speech and Signal
2310 Processing, pp 1461–1464,
2311 Torok MM, Golparvar-Fard M, Kochersberger KB (2014) Image-based automated 3D crack
2312 detection for post-disaster building assessment. Journal of Computing in Civil Engineer-
2313 ing 28:A4014004
2314 Trafalis TB, Adrianto I, Richman MB, Lakshmivarahan S (2014) Machine-learning classifiers
2315 for imbalanced tornado data. Computational Management Science 11:403–418
2316 Trugman DT, Shearer PM (2017) GrowClust: A hierarchical clustering algorithm for relative
2317 earthquake relocation, with application to the Spanish Springs and Sheldon, Nevada,
2318 earthquake sequences. Seismological Research Letters 88:379–391
2319 Tucker I, Gil-Garcia JR, Sayogo DS (2017) Collaborative data analytics for emergency re-
2320 sponse: Identifying key factors and proposing a preliminary framework. In: Proceedings
2321 of the 10th International Conference on Theory and Practice of Electronic Governance,
2322 pp 508–515
2323 Tunusluoglu MC, Gokceoglu C, Nefeslioglu HA, Sonmez H (2007) Extraction of potential
2324 debris source areas by logistic regression technique: a case study from Barla, Besparmak
2325 and Kapi mountains (NW Taurids, Turkey). Environmental Geology 54:9–22
2326 Uddin MN, Islam AKMS, Bala SK, Islam GMT, Adhikary S, Saha D, Haque S, Fahad MGR,
2327 Akter R (2019) Mapping of climate vulnerability of the coastal region of Bangladesh
2328 using principal component analysis. Applied Geology 102:47–57
2329 Ukkusuri SV, Zhan X, Sadri AM, Ye Q (2014) Use of social media data to explore crisis in-
2330 formatics: Study of 2013 Oklahoma Tornado. Transportation Research Record: Journal
2331 of the Transportation Research Board (2459):110–118
2332 UN (2015) Sendai framework for disaster risk reduction: 2015 - 2030. URL https://www.
2333 unisdr.org/files/43291 sendaiframeworkfordrren.pdf
2334 USGS (2016a) ShakeCast. URL https://earthquake.usgs.gov/research/software/shakecast.
2335 php
2336 USGS (2016b) ShakeMap. URL https://earthquake.usgs.gov/data/shakemap/
2337 S̆ubik S, Rohde S, Weber T, Wietfeld C (2010) SPIDER: Enabling interoperable informa-
2338 tion sharing between public institutions for efficient disaster recovery and response. In:
2339 Proceedings of the 2010 IEEE International Conference on Technologies for Homeland
2340 Security, IEEE, pp 190–196, DOI: 10.1109/THS.2010.5655061
2341 Uunk L, Wijinberg K, Morelissen R (2010) Automated mapping of the intertidal beach
2342 bathymetry from video images. Ocean Engineering 57:461–469
2343 Valkaniotis S, Papathanassiou G, Ganas A (2018) Mapping an earthquake-induced landslide
2344 based on UAV imagery: case study of the 2015 Okeanos landslide, Lefkada, Greece.
2345 Engineering Geology 245:141–152
2346 van Aardt JA, McKeown D, Faulring J, Raqueño N, Casterline M, Renschler C, Eguchi R,
2347 Messinger D, Krzaczek R, Cavillia S, Antalovich Jr J, Philips N, Bartlett B, Salvaggio
2348 C, Ontiveros E, Gill S (2011) Geospatial disaster response during the Haiti Earthquake:
2349 A case study spanning airborne deployment, data collection, transfer, processing, and
2350 dissemination. Photogrammetric Engineering & Remote Sensing 9:943–952
2351 Vanschoren J, van Rijn JN, Bischl B, Torgo L (2014) OpenML: networked science in machine
2352 learning. ACM SIGKDD Explorations Newsletter 15:49–60
 54 Wenjuan Sun et al.

2353 Velev D, Zlateva P, Zong X (2018) Challenges of 5G usability in disaster management. In:
2354 Proceedings of the 2018 International Conference on Computing and Artificial Intelli-
2355 gence (ICCAI 2018), ACM, pp 71–75
2356 Verma S, Vieweg S, Corvey WJ, Palen L, Martin JH, PAlmer M, Anderson KM (2011)
2357 Natural language processing to the rescue? extracting “situational awareness” tweets
2358 during mass emergency. In: Proceedings of the fifth international conference on Weblogs
2359 and Social Media (ICWSM2011), AAAI, pp 385–392, URL https://www.aaai.org/ocs/
2360 index.php/ICWSM/ICWSM11/paper/view/2834
2361 Verma S, Nair HS, Agarwal G, Dhar J, Shukla A (2020) Deep reinforcement learning for
2362 single-shot diagnosis and adaptation in damaged robots. In: Proceedings of the 7th ACM
2363 IKDD CoDS and 25th COMAD, ACM, pp 82–89
2364 Vetrivel A, Kerle N, Gerke M, Nex G, Vosselman G (2016) Towards automated satellite
2365 image segmentation and classification for assessing disaster damage using data-specific
2366 features with incremental learning. In: Kerle N, Gerke M, Lefevre S (eds) Proceedings
2367 of the GEOBIA2016: Solutions and Synergies, Enschede: University of Twente, Faculty
2368 of Geo-Information Science and Earth Observation (ITC), pp 1–5, DOI: 10.3990/2.369
2369 Vetrivel A, Gerke M, Kerle N, Nex F, Vosselman G (2018) Disaster damage detection through
2370 synergistic use of deep learning and 3D point cloud features derived from very high
2371 resolution oblique aerial images, and multiple-kernel-learning. ISPRS Journal of Pho-
2372 togrammetry and Remote Sensing 140:45–59
2373 Victores JG, Martı́nez S, Balaguer AJC (2011) Robot-aided tunnel inspection and mainte-
2374 nance system by vision and proximity sensor integration. Automation in Construction
2375 20:629–636
2376 Vieweg S (2012) Situational awareness in mass emergency: A behavioral and linguistic anal-
2377 ysis of microblogged communications. PhD dissertation, University of Colorado at Boul-
2378 der
2379 Wahab AM, Ludin ANM (2018) Flood vulnerability assessment using artificial neural net-
2380 works in Muar Region, Johor Malaysia. IOP Conference Series: Earth and Environmen-
2381 tal Science 169:012056
2382 Wan C, Mita A (2010) Early warning of hazard for pipelines by acoustic recognition using
2383 principal component analysis and one-class support vector machines. Smart Structures
2384 and Systems 17:405–421
2385 Wang B, Li Y, Ming W, Wang S (2020a) Deep reinforcement learning method for demand
2386 response management of interruptible load. IEEE Transactions on Smart Grid DOI:
2387 10.1109/TSG.2020.2967430
2388 Wang C, Wu J, Wang X, He X (2018a) Application of the hidden Markov model in a dynamic
2389 risk assessment of rainstorms in Dalian, China. Stochastic Environmental Research and
2390 Risk Assessment 32:2045–2056
2391 Wang F, Jin Z, Qian X, Ren Q, Huo Z (2012) Fuzzy clustering analysis in determining sub-
2392 faults of large earthquakes based on aftershock distribution. Acta Seismologica Sinica
2393 34:793–803, in Chinese
2394 Wang J, Zhang W, Ip W (2010a) An integrated road construction and resource planning
2395 approach to the evacuation of victims from single source to multiple destinations. IEEE
2396 Transactions on Intelligent Transportation Systems 11:277–289
2397 Wang J, Zhu S, Gong Y (2010b) Driving safety monitoring using semi-supervised learning on
2398 time series data. IEEE Transactions on Intelligent Transportation Systems 11:728–737
2399 Wang J, Wu Y, Yen N, Guo S, Cheng Z (2016) Big data analytics for emergency communi-
2400 cation networks: a survey. IEEE Communications Surveys & Tutorials 18:1758–1778
2401 Wang J, Zhang J, Gong L, Li Q, Zhou D (2018b) Indirect seismic economic loss assess-
2402 ment and recovery evaluation using nighttime light images application for Wenchuan
2403 earthquake. Natural Hazards and Earth System Sciences 18:3253–3266
2404 Wang JH, Lin GF, Chang MJ, Huang IH, Chen YR (2019a) Real-time water-level forecast-
2405 ing using dilated causal convolutional neural networks. Water Resources Management
2406 33:3759–3780
2407 Wang JP, Huang D, Chang SC, Brant L (2013a) On-site earthquake early warning with mul-
2408 tiple regression analysis: Featuring two user-friendly applications for Excel. Computers
2409 & Geosciences 58:1–7
 Applications of artificial intelligence for disaster management 55

2410 Wang K, Shi X, Goh APX, Qian S (2019b) A machine learning based study on pedestrian
2411 movement dynamics under emergency evacuation. Fire Safety Journal 106:163–176
2412 Wang L, Sawada K, Moriguchi S (2013b) Landslide susceptibility analysis with logistic
2413 regression model based on FCM sampling strategy. Computers & Geosciences 57:81–92
2414 Wang L, Abdel-Aty M, Lee J, Shi Q (2019c) Analysis of real-time crash risk for express-
2415 way ramps using traffic, geometric, trip generation, and socio-demographic predictors.
2416 Accident Analysis and Prevention 122:378–384
2417 Wang LC, Lai CC, Shuai HH, Lin HP, Li CY, Cheng TH, Chen CH (2019d) Communica-
2418 tions and networking technologies for intelligent drone cruisers. In: Proceedings of the
2419 2019 IEEE Globecom Workshops on Space-Ground Integrated Networks, IEEE, DOI:
2420 10.1109/GCWkshps45667.2019.9024679
2421 Wang P, Tan E, Jin Y, Wang J, Wang L (2019e) A deep reinforcement leaning evolution of
2422 emergency state during traffic network. In: Proceedings of the 14th IEEE Conference
2423 on Industrial Electronics and Applications, IEEE, DOI: 10.1109/ICIEA.2019.8833977
2424 Wang Q, Guo Y, Yu L, Li P (2020b) Earthquake prediction based on spatio-temporal
2425 data mining: An LSTM network approach. IEEE Transactions on Emerging Topics in
2426 Computing 8:148–158, DOI: 10.1109/TETC.2017.2699169
2427 Wang W (2018) Human detection based on radar sensor network in natural disaster. In:
2428 Geological Disaster Monitoring Based on Sensor Networks, Springer, pp 109–134
2429 Wang Y, Wang Q, Shi S, He X, Tang Z, Zhao K, Chu X (2019f) Benchmarking the perfor-
2430 mance and power of AI accelerators for AI training. URL https://arxiv.org/abs/1910.
2431 10045
2432 Wang Z, Wu J, Cheng L, Liu K, Wei Y (2018c) Regional flood risk assessment via coupled
2433 fuzzy c-means clustering methods: an empirical analysis from China’s Huaihe River
2434 Basin. Natural Hazards 93:803–822
2435 Wang Z, Lam NS, Obradovich N, Ye X (2019g) Are vulnerable communities digitally left
2436 behind in social responses to natural disasters? an evidence from Hurricane Sandy with
2437 Twitter data. Applied Geography 108:1–8
2438 Wang ZN, Chen J, Chen WC, Arulrajah A, Horpibulsuk S (2018d) Investigation into the
2439 tempo-spatial distribution of recent fire hazards in China. Natural Hazards 92:1889–1907
2440 Wanik DW, Anagnostou EN, Hartman BM, Frediani MEB, Astitha M (2015) Storm outage
2441 modeling for an electric distribution network in Northeastern USA. Natural Hazards
2442 79:1359–1384
2443 Watson D, Clark L (1994) The PANAS-X: Manual for the positive and negative affect
2444 schedule - expanded form. Tech. rep., The University of Iowa, https://ntrs.nasa.gov/
2445 search.jsp?R=20150003831
2446 Wilson R, zu Erbach-Schoenberg E, Albert M, Power D, Tudge S, Gonzalez M, Guthrie S,
2447 Chamberlain H, Brooks C, Hughes C, Pitonakova L, Buckee C, Lu X, Wetter E, Tatem
2448 A, Bengtsson L (2016) Rapid and near real-time assessments of population displacement
2449 using mobile phone data following disasters: The 2015 Nepal earthquake. PLoS Currents
2450 Disasters 24, DOI: 10.1371/currents.dis.d073fbece328e4c39087bc086d694b5c
2451 Wilts A (2018) Natural disaster damage cost America $306 bil-
2452 lion in 2017. https://www.independent.co.uk/news/world/americas/
2453 natural-disasters-us-damage-cost-money-2017-a8148771.html
2454 Wlwood E, Corotis RB (2015) Application of fuzzy pattern recognition of seismic damage
2455 to concrete structures. ASCE-ASME Journal of Risk and Uncertainty in Engineering
2456 Systems, Part A: Civil Engineering 1:04015011
2457 Wu D, Yan D, Yang G, Wang X, Xiao W, Zhang H (2013) Assessment on agricultural
2458 drought vulnerability in the Yellow River basin based on a fuzzy clustering iterative
2459 model. Natural Hazards 67:919–936
2460 Wu K, Yang S, Zhu KQ (2015) False rumors detection on sina weibo by propagation struc-
2461 tures. In: The IEEE 31st International Conference on Data Engineering, IEEE, pp 651–
2462 662
2463 Wu Q, Xu H, Pang W (2008) GIS and ANN coupling model: an innovative approach to
2464 evaluate vulnerability of karst water inrush in coalmines of north China. Environmental
2465 Geology 54:937–943
2466 Xiao Y, Li B, Gong Z (2018) Real-time identification of urban rainstorm waterlogging dis-
2467 asters based on Weibo big data. Natural Hazards 94:833–842
 56 Wenjuan Sun et al.

2468 Xiong J, Li J, Cheng W, Wang N, Guo L (2019) A gis-based support vector machine model
2469 for flash flood vulnerability assessment and mapping in china. International Journal of
2470 Geo-Information 8:297
2471 Xu C, Dai F, Xu X, Yuan, Lee H (2012) GIS-based support vector machine modeling of
2472 earthquake-triggered landslide susceptibility in the Jianjiang River watershed, China.
2473 Geomorphology 145–146:70–80
2474 Xu JX, Lu W, Li Z, Khaitan P, Zaytseva V (2019a) Building damage detection in satellite
2475 imagery using convolutional neural networks. In: Proceedings of the 33rd Conference on
2476 Neural Information Processing Systems (NeurIPS 2019), arXiv:1910.06444
2477 Xu N, Guikema SD, Davidson RA, Nozick LK, Çağnan Z, Vaziri K (2007) Optimizing
2478 scheduling of post-earthquake electric power restoration tasks. Earthquake Engineering
2479 & Structural Dynamics 36:265–284
2480 Xu N, Zhang Q, Zhang H, Hong M, Akerkard R, Liang Y (2019b) Global optimization
2481 for multi-stage construction of rescue units in disaster response. Sustainable Cities and
2482 Society 51:101768
2483 Yabe T, Ukkusuri SV (2019) Integrating information from heterogeneous networks on social
2484 media to predict post-disaster returning behavior. Journal of Computational Science
2485 32:12–20
2486 Yadollahnejad V, Bozorgi-Amiri A, Jabalameli M (2017) Allocation and vehicle routing for
2487 evacuation operations: A model and a simulated annealing heuristic. Journal of Urban
2488 Planning and Development 143:04017018
2489 Yamaguchi K, Shirota Y (2019) Pattern classification of disasters impact on companies stock
2490 prices. In: The 10th International Conference on Awareness Science and Technology
2491 (iCAST), IEEE, DOI: 10.1109/ICAwST.2019.8923362
2492 Yan J, He H, Zhong X, Tang Y (2016) Q-learning-based vulnerability analysis of smart grid
2493 against sequential topology attacks. IEEE Transactions on Information Forensics and
2494 Security 12:200–210
2495 Yan Y, Kong L, He X, Ouyang M, Peeta S, Chen X (2017) Pre-disaster investment deci-
2496 sions for strengthening the Chinese railway system under earthquakes. Transportation
2497 Research Part E: Logistics and Transportation Review 105:39–59
2498 Yang B, Liu M (2018) Keeping in touch with collaborative UAVs: a deep reinforcement learn-
2499 ing approach. In: Proceedings of the Twenty-Seventh International Joint Conference on
2500 Artificial Intelligence (IJCAI-18), pp 562–568
2501 Yang J, Yu M (2011) A model for seismic vulnerability score assignment of road infrastruc-
2502 ture using linear regression technique. Applied Mechanics and Materials 147:266–269
2503 Yang J, Chesbrough H, Hurmelinna-Laukkanen P (2019a) The rise, fall, and resurrection
2504 of IBM Watson Health. Tech. rep., Haas School of Business, University of California,
2505 http://jultika.oulu.fi/files/nbnfi-fe2020050424858.pdf
2506 Yang L, Cervone G (2019) Analysis of remote sensing imagery for disaster assessment using
2507 deep learning: a case study of flooding event. Soft Computing 23:13393–13408
2508 Yang S, Yang D, Chen J, Zhao B (2019b) Real-time reservoir operation using recurrent
2509 neural networks and inflow forecast from a distributed hydrological model. Journal of
2510 Hydrology 579:142229
2511 Yang T, Xie J, Li G, Mou N, Li Z, Tian C, Zhao J (2019c) Social media big data mining and
2512 spatio-temporal analysis on public emotions for disaster mitigation. ISPRS International
2513 Journal of Geo-Information 8:29
2514 Yang Y, Lu W, Domack J, Li T, Chen SC, Luis S, Navlakha JK (2012) MADIS:
2515 A multimedia-aided disaster information integration system for emergency manage-
2516 ment. In: Proceedings of the 8th International Conference on Collaborative Com-
2517 puting: Networking, Applications and Worksharing (CollaborateCom), IEEE, DOI:
2518 10.4108/icst.collaboratecom.2012.250525
2519 Yang Y, Chen G, Reniers G (2019d) Vulnerability assessment of atmospheric storage tanks to
2520 floods based on logistic regression. Reliability Engineering & System Safety 196:106721
2521 Yao Z, Zhang G, Lu D, Liu H (2019) Data-driven crowd evacuation: A reinforcement learning
2522 method. Neurocomputing 366:314–327
2523 Yaseen ZM, El-shafie A, Jaafar O, Afan HA, Sayla KN (2015) Artificial intelligence based
2524 models for stream-flow forecasting: 2000-2015. Journal of Hydrology 530:829 – 844
 Applications of artificial intelligence for disaster management 57

2525 Yeum CM, Dyke SJ (2015) Vision–based automated crack detection for bridge inspection.
2526 Computer-Aided Civil and Infrastructure Engineering 30:759–770
2527 Yilmaz I (2010) Comparison of landslide susceptibility mapping methodologies for Koyul-
2528 hisar, Turkey: conditional probability, logistic regression, artificial neural networks, and
2529 support vector machine. Environmental Earth Sciences 61:821–836
2530 Yin J, Karimi S, Robinson B, Cameron M (2012a) ESA: Emergency situation awareness via
2531 microbloggers. In: The 21st ACM International Conference on Information and Knowl-
2532 edge Management, ACM, pp 2701–2703
2533 Yin J, Lampert A, Cameron M, Robinson B, Power R (2012b) Using social media to enhance
2534 emergency situation awareness. IEEE Intelligent Systems 27:52 – 59
2535 Yoon DK, Jeong S (2016) Assessment of community vulnerability to natural disasters in Ko-
2536 rea by using GIS and machine learning techniques. In: Quantitative Regional Economic
2537 and Environmental Analysis for Sustainability in Korea, Springer, pp 123–140
2538 Yoon H, Shiftehfar R, Cho S, Spencer BFJ, Nelson ME, Agha G (2016) Victim localiza-
2539 tion and assessment system for emergency responders. Journal of Computing in Civil
2540 Engineering 30:04015011
2541 Yu L, Zhu JH (2014) Nonlinear damage detection using higher statistical moments
2542 of structural responses. Structural Engineering and Mechanics 54(2):221–237, DOI:
2543 10.12989/sem.2015.54.2.221
2544 Yu L, Wang N, Meng X (2005) Real-time forest fire detection with wireless sensor networks.
2545 In: International Conference on Wireless Communications, Networking and Mobile Com-
2546 puting, IEEE, pp 1214–1217
2547 Yu L, Zhu JH, Yu LL (2016) Structural damage detection in a truss bridge model using
2548 fuzzy clustering and measured FRF data reduced by principal component projection.
2549 Advances in Structural Engineering 16:207–217
2550 Yu M, Yang C, Li Y (2018) Big data in natural disaster management: a review. Geosciences
2551 8:165
2552 Yu M, Huang Q, Qin H, Scheele C, Yang C (2019) Deep learning for real-time social media
2553 text classification for situation awareness using Hurricanes Sandy, Harvey, and Irma as
2554 case studies. International Journal of Digital Earth 12:1230–1247
2555 Yu PS, Yang TC, Chen SY, Kuo CM, Tseng HW (2017) Comparison of random forests
2556 and support vector machine for real-time radar-derived rainfall forecasting. Journal of
2557 Hydrology 552:92–104
2558 Yu S, Kim SW, Oh CW, An H, Kim JM (2014) Quantitative assessment of disaster resilience:
2559 An empirical study on the importance of post-disaster recovery costs. Reliability Engi-
2560 neering & System Safety 137:6–17
2561 Yuan C, Moayedi H (2019) Evaluation and comparison of the advanced metaheuristic and
2562 conventional machine learning methods for the prediction of landslide occurrence. En-
2563 gineering with Computers DOI: 10.1007/s00366-019-00798-x
2564 Yuan C, Liu Z, Liu Z (2015) A survey on technologies for automatic forest fire monitoring,
2565 detection, and fighting using unmanned aerial vehicles and remote sensing techniques.
2566 Canadian Journal of Forrest Research 45:783–792
2567 Zagorecki AT, Johnson DE, Ristvej J (2013) Data mining and machine learning in the con-
2568 text of disaster and crisis management. International Journal of Emergency Management
2569 9:351–365
2570 Zahran S, Brody SD, Peacock WG, Vedlitz A, Grover H (2008) Social vulnerability and the
2571 natural and built environment: a model of flood casualties in Texas. Disasters 32:537–560
2572 Zare M, Pourghasemi HR, Vafakhah M, Pradhan B (2013) Landslide susceptibility map-
2573 ping at Vaz Watershed (Iran) using an artificial neural network model: a comparison
2574 between multilayer perceptron (MLP) and radial basic function (RBF) algorithms. Ara-
2575 bian Journal of Geosciences 6:2873–2888
2576 Zeng Y, Yan Y, Weng S, Sun Y, Tian W, Yu H (2018) Fuzzy clustering of time-series model
2577 to damage identification of structures. Advances in Structural Engineering 22:868–881
2578 Zhang D, Zhang Y, Li Q, Plummer T, Wang D (2019a) CrowdLearn: A crowd-AI hy-
2579 brid system for deep learning-based damage assessment applications. In: Proceedings of
2580 the IEEE 39th International Conference on Distributed Computing Systems (ICDCS),
2581 IEEE, pp 1221–1232, DOI: 10.1109/ICDCS.2019.00123
 58 Wenjuan Sun et al.

2582 Zhang G, Li B, Li Z, Wang C, Zhang H, Shang H, Hu W, Zhang T (2014) De-

2583 velopment of robotic spreader for earthquake rescue. In: Proceedings of the 2014
2584 IEEE International Symposium on Safety, Security, and Rescue Robotics, IEEE, DOI:
2585 10.1109/SSRR.2014.7017679
2586 Zhang H (2016) Household vulnerability and economic status during disaster recovery and its
2587 determinants: a case study after the Wenchuan earthquake. Natural Hazards 83:1505–
2588 1526
2589 Zhang H, Li N, Zhang W, Pei X (2016a) Experiments to automatically monitor drought
2590 variation using simulated annealing algorithm. Natural Hazards 84:175–184
2591 Zhang J (2004) Risk assessment of drought disaster in the maize-growing region of Songliao
2592 Plain, China. Agriculture, Ecosystems & Environment 102:133–153
2593 Zhang K, Niroui F, Fococelli M, Nejat G (2018a) Robot navigation of environments with
2594 unknown rough terrain using deep reinforcement learning. In: Proceedings of the 2018
2595 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR), DOI:
2596 10.1109/SSRR.2018.8468643
2597 Zhang L, Zhang L, Du B (2016b) Deep learning for remote sensing data: A technical tutorial
2598 on the state of the art. IEEE Geoscience and Remote Sensing Magazine 4:22–40
2599 Zhang L, Lv X, Dhakal S (2019b) A reinforcement learning-based stakeholder value aggre-
2600 gation model for collaborative decision making on disaster resilience. In: ASCE Inter-
2601 national Conference on Computing in Civil Engineering 2019, ASCE, pp 490–497
2602 Zhang R, Rezaee Z, Zhu J (2010) Corporate philanthropic disaster response and ownership
2603 type: Evidence from Chinese firms response to the Sichuan Earthquake. Journal of
2604 Business Ethics 91:51–63, article No. 51
2605 Zhang Y, Burton HV (2019) Pattern recognition approach to assess the residual structural
2606 capacity of damaged tall buildings. Structural Safety 78:12–22
2607 Zhang Y, Peacock WG (2009) Planning for housing recovery? lessons learned from Hurricane
2608 Andrew. Journal of the American Planning Association 76:5–24
2609 Zhang Y, Burton HV, Sun H, Shokrabadi M (2018b) A machine learning framework for
2610 assessing post-earthquake structural safety. Structural Safety 72:1–16
2611 Zhao J, Zhu L, Liu GF, Liu G, Han Z (2009) A modified genetic algorithm for global
2612 path planning of searching robot in mine disasters. In: 2009 International Conference on
2613 Mechatronics and Automation, IEEE, pp 4936–4940, DOI: 10.1109/ICMA.2009.5246026
2614 Zhao L, Hicks FE, Fayek AR (2012) Applicability of multilayer feed-forward neural networks
2615 to model the onset of river breakup. Cold Regions Science and Technology 70:32–42
2616 Zhao Q, Chen Z, Liu C, Luo N (2019) Extracting and classifying typhoon disaster in-
2617 formation based on volunteered geographic information from chinese sina microblog.
2618 Concurrency and Computation: Practice and Experience 31:e4910
2619 Zhao Y, Zheng Z, Zhang X, Liu Y (2017) Q-learning algorithm based UAV path learning and
2620 obstacle avoidance approach. In: Proceedings of the 36th Chinese Control Conference,
2621 DOI: 10.23919/ChiCC.2017.8027884
2622 Zhao Y, Meng X, Qi T, Qing F, Xiong M, Li Y, Guo P, Chen G (2020) AI-based iden-
2623 tification of low-frequency debris flow catchments in the Bailong River basin, China.
2624 Geomorphology 359:107125
2625 Zhao Z, Wu Y (2016) Attention-based convolutional neural networks for sentence classi-
2626 fication. In: The 17th Annual Conference of the International Speech Communication
2627 Association, pp 705–709
2628 Zheng L, Shen C, Tang L, Li T, Luis S, Chen SC (2011) Applying data mining techniques to
2629 address disaster information management challenges on mobile devices. In: Proceedings
2630 of the 17th ACM SIGKDD international conference on Knowledge discovery and data
2631 mining, ACM, pp 283–291
2632 Zheng L, Shen C, Tang L, Zeng C, Li T, Luis S, , Chen SC (2013a) Data mining meets
2633 the needs of disaster information management. IEEE Transactions on Human-Machine
2634 Systems 43:451–464
2635 Zheng S, Liu H (2019) Improved multi-agent deep deterministic policy gradient for path
2636 planning-based crowd simulation. IEEE Access 7:147755–47770
2637 Zheng YJ, Ling HF, Xue JY, Chen SY (2013b) Population classification in fire evacuation:
2638 A multiobjective particle swarm optimization approach. IEEE Transactions on Evolu-
2639 tionary Computation 18:70–81
 Applications of artificial intelligence for disaster management 59

2640 Zhong L, Garlichs K, Yamada S, Takano K, Ji Y (2018) Mission planning for UAV-based
2641 opportunistic disaster recovery networks. In: 2018 15th IEEE Annual Consumer Commu-
2642 nications & Networking Conference (CCNC), IEEE, DOI: 10.1109/CCNC.2018.8319233
2643 Zhou C, Yin K, Cao Y, Ahmed B, Li Y, Catani F, Pourghasemi HR (2018a) Landslide
2644 susceptibility modeling applying machine learning methods: A case study from Longju
2645 in the Three Gorges reservoir area, China. Computers & Geosciences 112:23–37
2646 Zhou J, Pei H, Wu H (2017a) Early warning of human crowds based on query data from
2647 Baidu maps: Analysis based on Shanghai Stampede. In: Big Data Support of Urban
2648 Planning and Management, Springer, pp 19–41
2649 Zhou YL, Maia NMM, Sampaio RPC (2017b) Structural damage detection using transmis-
2650 sibility together with hierarchical clustering analysis and similarity measure. Structural
2651 Health Monitoring 16:711–731, DOI: 10.1016/j.cag.2013.10.008
2652 Zhou YL, Maia NMM, Wahab MA (2018b) Damage detection using transmissibility com-
2653 pressed by principal component analysis enhanced with distance measure. Journal of
2654 Vibration and Control 24:2001–2019
2655 Zhu C, Wu J (2013) Hybrid of genetic algorithm and simulated annealing for support vector
2656 regression optimization in rainfall forecasting. International Journal of Computational
2657 Intelligence and Applications 12:1350012
2658 ZIDI I, Al-Omani M, Aldhafeeri K (2019) A new approach based on the hybridization of
2659 simulated annealing algorithm and Tabu search to solve the static ambulance routing
2660 problem. Procedia Computer Science 159:1216–1228
2661 Zobel CW (2014) Quantitatively representing nonlinear disaster recovery. Decision Sciences
2662 45:1053–1082
2663 Zorn CR, Shamsedin AY (2015) Post-disaster infrastructure restoration: a comparison of
2664 events for future planning. International Journal of Disaster Risk Reduction 13:158–166
2665 Zou L, Lam NSN, Shams S, Cai H, Meyer MA, Yang S (2019) Social and geographical
2666 disparities in Twitter use during Hurricane Harvey. International Journal of Digital
2667 Earth 12:1500–1318
2668 Zubiaga A, Liakata M, Procter R, Hoi GWS, Tolmie P (2016) Analysing how people orient
2669 to and spread rumours in social media by looking at conversational threads. PLoS ONE
2670 11:1–29
2671 Zubiaga A, Aker A, Bontcheva K, Liakata M, Procter R (2018) Detection and resolution of
2672 rumours in social media: A survey. ACM Computing Surveys 51:32