Hindawi

Complexity
Volume 2018, Article ID 1890643, 9 pages
https://doi.org/10.1155/2018/1890643

Research Article
The Spreading of Information in Online Social Networks
through Cellular Automata

Yuda Wang and Gang Li

School of Economics and Management, Beijing University of Posts and Telecommunications, Beijing 100876, China

Correspondence should be addressed to Yuda Wang; wangyd@bupt.edu.cn

Received 10 September 2018; Accepted 15 October 2018; Published 1 November 2018

Academic Editor: Dimitri Volchenkov

Copyright © 2018 Yuda Wang and Gang Li. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.

Epidemic dynamics in complex networks have been extensively studied. Due to the similarity between information and disease
spreading, most studies on information dynamics use epidemic models and merely consider the characteristics of online social
networks and individual’s cognitive. In this paper, we propose an online social networks information spreading (OSIS) model
combining epidemic models and individual’s cognitive psychology. Then we design a cellular automata (CA) method to provide a
computational method for OSIS. Finally, we use OSIS and CA to simulate the spreading and evolution of information in online social
networks. The experimental results indicate that OSIS is effective. Firstly, individual’s cognition affects online information spreading.
When infection rate is low, it prevents the spreading, whereas when infection rate is sufficiently high, it promotes transmission.
Secondly, the explosion of online social network scale and the convenience of we-media greatly increase the ability of information
dissemination. Lastly, the demise of information is affected by both time and heat decay rather than probability. We believe that these
findings are in the right direction for perceiving information spreading in online social networks and useful for public management
policymakers seeking to design efficient programs.

1. Introduction susceptible-exposed-infected-removed (SEIR) model [11–15].

With the growth of the Internet scale and the convenience
Networks such as Internet, communication networks, and of we-media, information such as rumors’ spreading and
social networks can be used to describe the interconnections evolution in online social networks has received special atten-
among individuals. Since most of these networks have non- tion [16, 17]. Due to the similarity between information and
trivial structural properties, they are called complex networks disease, the models mentioned above are widely considered.
[1–4]. According to the degree of each network node, complex Qian [18] investigated the spreading and evolution of
networks are divided into regular networks, Erdos-Renyi public opinion on Weibo [19] with SIR model. Xiong et
(ER) networks [5], small-world networks [6], and scale-free al. [20] proposed a new model called SCIR to describe the
(SF) networks [7]. Among those networks, the small-world spreading of information. Centola et al. [21] have illustrated
model proposed by Watts and Strogatz describes the features the effects of network structure on diffusion experimentally
of high clustering and small average path length, which is by studying the spread of health behavior through artificially
most suitable for real social networks. These small-world structured online communities. However, all these studies are
features [8] have been found to have a significant impact on based on a probabilistic approach, regardless of individual’s
the dynamics in networks [9]. cognitive psychology and the characteristics of online social
Spreading [10] is one of the dynamical processes on com- networks. Recently, researchers start to seriously take into
plex networks, especially the spreading of epidemic, which consideration the specific features of online social networks
has been investigated for decades and achieves a lot of the- information spreading, such as individual’s characteristics
ories. Most of these studies investigate susceptible-infected- and the impact of friends around [22–31]. Nevertheless, there
susceptible (SIS), susceptible-infected-recovered (SIR), or are more factors affecting online information spreading.
2 Complexity

Furthermore, most previous studies of dynamical pro- (iii) revealing the characteristic of online social networks
cesses on complex networks are based on mean field method information dissemination.
[32] or analytical equation [33]. However, these approaches
has some serious drawbacks. The method based on mean The sections in this paper are structured as follows: Section 1
field can only reflect an approximate trend from macroscopic presents the background while Section 2 describes the model;
view, which cannot accurately describe the spreading state Section 3 shows the design of cellular automata followed by
of each individual. Meanwhile, analytical equation neglects the experimental results in Section 4; finally, we conclude our
variable susceptibility of individuals and cannot handle the paper in Section 5 with the future work discussed.
different boundary and initial condition; in addition, it will
be very complicated for large-scale online social networks. 2. Online Social Network Information
Cellular automata (CA) [34, 35] proposed by Von Neumann Spreading Model
can overcome the above drawbacks and have been used by
several researchers as an alternative method of modeling 2.1. Analysis of Information in Online Social Network. Com-
epidemics. In this model every cell of the grid represents pared with traditional social networks, online social networks
an individual, updating state of each cell by state transition use Internet as the medium of communication rather than
criterion [36, 37]. face-to-face, which makes information transmission more
In summary, complex networks provide theoretical basis easy to forward. Meanwhile, in social networks such as Twit-
for social network topology, and the extensive investigation ter, information is transmitted synchronously to all friends,
of epidemic models promotes the quantitative study of which is different from the peer-to-peer communication
information spreading. By referring to complex network of offline networks. The communication of information is
dynamics, as well as combining the features of public opinion, more highly timely. Meanwhile, because of the intelligence of
a breakthrough has been made in the research on information terminals, the carrier of online information varies from text
transmission in social networks. With the rapid development to sound, picture, video, etc. The content becomes interesting
of mobile Internet and Internet of things, the online social and easy to understand. Based on the features described
network is fundamentally different from the traditional social above, we consider the information spreading model from
network. The spreading and control of information, espe- three aspects.
cially public opinion, has become a difficulty and key point
(i) Propagation factor: Due to the characteristics of
concerned by public management policymakers. Meanwhile,
online social networks and individual’s fast-food
due to the complexity of network topology and the explosion
reading habits, the cognition of information depends
of network access nodes, traditional analytical methods have
not only on the content itself, but also on the heat.
been unable to accurately calculate the results. By introducing
The more the friends who are concerned around a
the cellular automata theory, the bottleneck can be solved
topic, the higher the probability that an individual will
through the iterative calculation.
spread it.
In this paper, we propose an online social networks
information spreading (OSIS) model combining individual’s (ii) Memory superposition factor: Online social net-
cognitive psychology with traditional SEIR model. Moreover, works shorten the distance among people, which
we provide a computational method to prove the accuracy makes people access the same information over a
and authenticity of OSIS by cellular automata. We obtained continuous period of time; we call this phenomenon
experimental results with the help of computer graph theory. memory superposition. The memory superposition
The results reveal that individual’s cognition affects online enhances people’s curiosity, thus prompting the
information spreading. When infection rate is low, cognitive spreading.
psychology prevents the spreading. By contrast, when infec- (iii) Information attenuation factor: In traditional epi-
tion rate is sufficiently high, it promotes transmission. We demic models, a disease is cured via a certain prob-
also demonstrate that the explosion of online social network ability, whereas the demise of information is affected
scale and the convenience of we-media can greatly increase by both time and heat decay.
the ability of information dissemination, especially rumors.
Finally, we show that the demise of information is affected by 2.2. Description of OSIS Model. Since information spreading
the dual effects of the decay of time and heat. In particular, is similar to epidemic, we propose our online social network
the major contributions of our work can be summarized as information spreading (OSIS) model by combining the char-
follows: acteristics of online social networks mentioned above with
epidemic SEIR model. In OSIS, each individual adopts one of
(i) a new information spreading model that takes into four states: unaware, concerned, interested, and abandoned.
account individual’s cognitive psychology and the Unaware represents the individual who has not heard the
characteristics of online social networks: herd men- information and has the potential to accept it. Concerned
tality factor, memory superposition factor, and atten- stands for the person who is aware of the news but not
tion attenuation factor; willing to transmit it currently. Interested means the one
believes in the news and has the ability to spread it. And
(ii) a two-dimensional CA method that can accurately abandoned denotes the individual who loses interests and
analyze the process of online information spreading; never transmits the information again. We use complex
Complexity 3

Define:
concerned (1) State ←󳨀 current state of a node
(2) 𝜏 ←󳨀 time step
(3) Duration ←󳨀 the period in current state
Procedure:
unaware interested abandoned (4) for each t ∈ 𝜏 do
(5) Traverse the network and fill in State of each node
(6) for each node in small-world networks do
(7) if State = unaware then
Figure 1: The state transfer diagram. (8) if Propagation factor then
(9) State ←󳨀 interested
(10) else
(11) State ←󳨀 concerned
networks as the model of real social networks, in which
(12) end if
each node and edge present individual and online social (13) else if State = concerned then
relationship. (14) if Memory superposition factor then
As shown in Figure 1, the state of each node can be (15) State ←󳨀 interested
transformed over time. Initially, we randomly pick up one (16) else
node as an interested node and the rest are in unaware state. (17) Duration ←󳨀 Duration + 1
At each time step, an unaware state individual receives a topic (18) end if
from her neighbor and turns to interested or concerned state (19) else if State = interested then
according to the propagation factor. Furthermore, whether or (20) if Information attenuation factor then
not a concerned state node turns to interested state is based (21) State ←󳨀 abandoned
on memory superposition factor. Additionally, information (22) else
(23) Duration ←󳨀 Duration + 1
attenuation factor determines interested to abandoned state.
(24) end if
(25) end if
2.3. The Algorithm. Based on OSIS model described above, (26) end for
an algorithm for simulation of information spreading on (27) end for
complex networks has been developed. The core idea in
Algorithm 1 is presented with pseudocode. Algorithm 1: The information spreading model pseudocode.

(a) If a node is in unaware state and there are interested

state nodes around, propagation factor determines
is believed, which reduces the probability of spread-
state transition. In our model, the dissemination
ing. Similarly, we use the proportion of abandoned
of information is due to not only the information
individuals in the neighbors to calculate the impact of
itself, but also to herd mentality. For that reason, the
propagation factor consists of two parts. One part herd effect, and the expression is (1−𝑒−𝑛(𝑡)/𝑘 )/(1−1/𝑒).
is due to the information itself, here we adopt the Above all, the herd mentality effect can be described
algorithm in traditional SIR model. The infection rate as ((1−𝑒−𝑚(𝑡)/𝑘 )−(1−𝑒−𝑛(𝑡)/𝑘 ))/(1−1/𝑒).
is 𝛼 and the number of interested neighbors is m(t) In this paper the propagation factor is composed of
at the tth time. The node will have a probability of these two parts by the assigned 1/2 weight for the
1−(1−𝛼)𝑚(𝑡) being infected. The other part is caused expressiveness of the model and the computability of
by herd mentality, which includes both positive and simulations, which can be expressed in (1). The greater
negative effect. We assume that for a given individual p(t), the more possibility of transferring state from
at the tth time, she has k adjacent, m(t) interested, unaware to interested.
and n(t) abandoned neighbors. The proportion of
1 1
interested individuals in the neighbors determines 𝑝 (𝑡) = ∗ (1 − (1 − 𝛼)𝑚(𝑡) ) +
the positive influence of the herd effect. When no 2 2
one around propagates the information (i.e., 𝑚(𝑡)/𝑘 (1)
(1 − 𝑒−𝑚(𝑡)/𝑘 ) + (𝑒−𝑛(𝑡)/𝑘 − 1)
= 0), the influence probability of herd effect is 0. ∗
While all friends are in interested state (i.e., 𝑚(𝑡)/𝑘 1 − 1/𝑒
= 0), the influence of herd effect is the greatest, the
probability of which is 1. Our observations show that (b) If a node is in concerned state, memory superposition
the effect of this ratio is a positively exponential effect factor determines state transition. In our designed
rather than a linear effect; we define the effect as model, we believe that human memory has an impact
(1−𝑒−𝑚(𝑡)/𝑘 )/(1−1/𝑒). By contrast, if most friends are on the spread of information. It is easy for an
in abandoned state, this may reduce the transmissi- individual in concerned state to change their state
bility. The more the neighbors in abandoned state, the by constantly having neighbors to spread or reject
lower the probability that the truth of the information the message within a short period of time. When
4 Complexity

different friends forward the same topic multiple

times, it will promote the concerned state to become
the interested state. By contrast, if not heard for a
long time, the topic will be abandoned. We define
the one who receives the information over the time
as s(t), which quantitatively describes the cumulative
effect of the memory. Initially set s(t) to 0; i.e., the
variable is 0 when a node changes from an unaware
state to a concerned state for the first time at tth time
step. If there are new turned interested nodes around
the next time step, then s(t+1) = s(t) +1. Otherwise,
s(t+1) = s(t) - 1. We set threshold 𝜀. When memory
superposition factor accumulates above 𝜀, the state
turns to interested. Contrarily if s(t) < -𝜀, it becomes Figure 2: The neighborhood of each cell is formed by the cells
abandoned state. associated with edges.
(c) If the node is in interested state, information atten-
uation determines state transition. Attenuation is a
compound process; here we consider both the interest Memory superposition factor is able to determine
decay and time decay. We use exponential function to concerned node’s next state. Furthermore, informa-
represent time decay and the proportion of interested tion attenuation factor can turn interested to aban-
individuals in the neighbors to represent interest doned state. The detailed transfer relationship can be
decay. Thus information attenuation factor can be found in Figure 1.
expressed as (2). We set threshold 𝛿; when a(t) decays
below the threshold, the interested state node will 4. Experiments and Evaluation
turn to abandoned state.
4.1. Implementation of Online Social Network. We use small-
−𝑡 𝑚 (𝑡) world network to simulate online social network in our
𝑎 (𝑡) = 𝑒 ∗ (2)
𝑘 experiments. On the one hand, small-world theory is also
called six degrees of separation, which means the distance
3. Design of Cellular Automata between two strangers is at most six people. This feature is
Cellular automata (CA) constitute the model of physical the closest to the online social network. On the other hand,
systems, where space and time are discrete and interactions on the basis of sampling and analysis on millions of level
are local. It is an effective theoretical tool for studying data excavated from the social groups in [38], online social
complex systems and, at the same time, can be simulated network has small-world effect.
exactly by computer program because of their intrinsic The implementation of a small-world network comes
discreteness. For those reasons, we adopt CA to realize OSIS from a regular lattice with a linear size L and with N =
model. In CA the value of each cell presents the local state, L∗L nodes. The initial node degree is k, which means each
which can only interact with its neighbors and is incapable of node is connected to its k/2 nearest neighbors clockwise and
global communication. The state is updated simultaneously at counterclockwise, deleting one link randomly, adding a link
discrete time steps according to the states of their neighbors between two randomly chosen nodes with probability p, and
at the preceding time step. Here are the elements of our avoiding duplicate links and self-loops. The link probability
proposed cellular automata. p turns the nature of the network between that of a regular
network (p = 0) and that of a random network (p = 1).
(a) Cellular space: Our model is set up in complex
network; each node in the network consists of the 4.2. Experimental Results. In our experiments, we use C++
cellular space. programming language to build the small-world networks
(b) Cellular neighborhood: For the reason that our cellu- with 2500 nodes and an average degree of 6. The shortest dis-
lar space is not the regular two-dimensional network, tance between two nodes is calculated by Dijkstra’s algorithm
we cannot adopt traditional Von Neumann or Moore in computer graph theory. The results are averaged over 100
structure. The neighborhood in Figure 2 illustrates realizations.
that the nodes associated with edges can be regarded
as adjacent relationship, which we call neighbors. 4.2.1. The Comparison with Traditional SEIR Model. To
demonstrate the effectiveness of OSIS, we compare our
(c) Cellular state: According to OSIS model, the state model with traditional epidemics SEIR model. We firstly
of each cell is unaware, concerned, interested, or observe the information spreading over time and space,
abandoned. considering various infection rates. Afterwards, we illustrate
(d) State transfer rule: As the methodology described the evolution. Particularly, unaware, concerned, interested,
above, unaware state turns to whether interested or and abandoned state correspond to susceptible, exposed,
concerned state is decided by propagation factor. infected, and recovered, respectively, in SEIR.
Complexity 5

1.0 1.0

percentage 0.8 0.8

percentage
0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
t t
unaware interested unaware interested
concerned abandoned concerned abandoned
(a) (b)

1.0 1.0

0.8 0.8

percentage
percentage

0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
t t
unaware interested susceptible infected
concerned abandoned exposed recovered
(c) (d)

1.0 1.0

0.8 0.8
percentage
percentage

0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
t t
susceptible infected susceptible infected
exposed recovered exposed recovered
(e) (f)

Figure 3: The percentage of each state in a continuous period of time. The spreading model is OSIS ((a), (b), (c)) and SEIR ((d), (e), (f))
separately. The infection rate is 0.35 in (a) and (d), 0.55 in (b) and (f), 0.75 in (c) and (e).

Firstly, we set up three sets of contrastive experiments. infection rate is 0.35 and 0.55 in the first and second set, which
Without generality, we set the rewiring probability of the is 0.75 in the third one.
small-world networks 1, i.e., a “complete” small-world net- Initially, one node in the centre is infected. After 30 time
work. And the threshold of the memory superposition factor periods, the percentage of each state is recorded and the result
is 1 and the threshold of information attenuation is 0.5. The is shown in Figure 3.
6 Complexity

(a)

(b)

(c)

(d)

Figure 4: The information spreading evolution in small-world. The spreading model is OSIS ((a), (b)) and SEIR ((c), (d)) separately. (a) and
(c) present the evolution of interested nodes, (b) and (d) stand for abandoned nodes. The infection rate is 0.35. The time period is 0, 5, 10, 15.

It can be seen from Figures 3(a) and 3(d) that when to 0.57 from 0.33. This is due to the fact that herd mentality
infection rate is low, the percentage of nodes in final aban- and memory superposition effect promote the spreading of
doned state is 94% in SEIR, which decreases as much as information.
38% to 58% in OSIS. This is because when the number Furthermore, to discover the evolution of OSIS, we
of initial infected nodes as well as information spreading investigate the information spreading over space. In this
capacity is low, herd mentality inhibits the transmission of experiment, rewiring probability of the small-world network
information. The phenomenon explains herd mentality effect is that 0.1 and 1 nodes in the centre of the network are infected
on online social networks from the reverse. Figures 3(b) and initially. It can be seen from Figure 4 that the evolution of
3(e) illustrate the peak of interested nodes is significantly online social network information is a gradual process, while
delayed, which is from the 7th time step in SEIR to the 13th the disease transmission has the characteristics of sudden
time step in OSIS. The result reveals that both propagation outbreak and the one exposed to the disease eventually
and disappearance of a topic is gradual over time or with the develops antibodies. Further comparing Figures 4(a) and
attention of people around, which is different from epidemic. 4(c), it is evident that the disease with SEIR model spreads
Figures 3(c) and 3(f) demonstrate that when the infection rate very fast and suddenly dies out. By contrast, the information
is high, the number of infected nodes is increased by 72% spreads throughout the network smoothly. Meanwhile, the
Complexity 7

2500 2500

2000 2000
number of abandoned

number of abandoned
1500 1500

1000 1000

500 500

0 0
0 10 20 30 40 50 60 70 0 10 20 3040 50 60 70
t t
initially 1 interested initially 1 interested
initially 10 interested initially 10 interested
initially 100 interested initially 100 interested
(a) (b)
2500

2000
number of abandoned

1500

1000

500

0
0 10 20 3040 50 60 70
t
initially 1 interested
initially 10 interested
initially 100 interested
(c)

Figure 5: The influence of different initial infection ratio in small-word. The rewiring probability is 0, 0.1, 1 in (a), (b), (c), respectively. Each
experiment randomly selects 1, 10, 100 initial interested nodes.

contrast between Figures 4(b) and 4(d) shows that the leads to a faster spreading and a higher steady-state infection
abandoned nodes in SEIR are in the large area, while the density. When the rewiring probability is 0.1, the peak is
number of immunized nodes in OSIS is flat. The results are reached at the 15th time period, while it reduces to 10th time
consistent with the above conclusions. period when the rewiring probability is 1. All the results above
reveal that the more initial infected nodes, the easier it is to
4.2.2. The Influence of Different Initial Infection Ratio. In spread information. However, this effect tends to converge
order to prove the influence of initial infection ratio on as the rewiring probability increases. Additionally, with the
information spreading, we randomly select 1, 10, and 100 strengthening of the small-world features (i.e., the average
initial interested nodes. The rewiring probability in each set number of friends per person in social networks is greater),
of experiments is 0, 0.1, and 1, respectively, and the infection the information can spread farther and inhibited harder.
rate is 0.35. The threshold of the memory superposition factor
is 1 and the threshold of information attenuation is 0.5. The 4.2.3. The Effect of Memory Superposition. To explore the
abandoned state nodes are recorded. memory superposition effect on information spreading. We
It is observable from Figure 5(a) that the number of initial perform simulations to obtain the infection density in the
infected nodes plays a decisive role in spreading when small- steady state as a function of the effective infection rate 𝛼
world effect is not obvious (i.e., rewiring probability is 0). for various superposition thresholds. In this experiment, the
When the number of initial interested nodes is 10, the number rewiring probability of the network is 0.1 and the threshold of
of the eventually abandoned nodes is 8 times more than that information attenuation is 0.5.
of 1 initial interested node, while the multiple is 60 times We set up three sets of experiments. The infection rate
when the number of initial interested nodes is 100 and 1. As in each set is 0.1, 0.5, and 0.7; meanwhile, the superposition
shown in Figures 5(b) and 5(c), a higher rewiring probability threshold is 1, 3, and 5. Initially, one node in network is
8 Complexity

our research shows that the demise of information is affected

2500 by the dual effects of the decay of time and heat.
Our current work mainly concentrates on the spreading
number of abandoned

2000
model itself. As for the future work, we believe that the real
online social network is not a single small-world network.
1500
Due to the vigorous development of Web 2.0, online networks
such as Weibo, Twitter, and Facebook are intertwined. We
1000
deem the information spreading in interconnected small-
500
world networks as the promising and challenging future
work.
0
0.1 0.5 0.7
infection rate Data Availability
superposition threshold=1
superposition threshold=3 The data used to support the findings of this study are
superposition threshold=5 available from the corresponding author upon request.

Figure 6: The number of abandoned node under different superpo-

sition threshold. Conflicts of Interest
The authors declare that they have no conflicts of interest.

randomly chosen to be infected. After 100 time periods, the References

number of nodes in abandoned state is recorded. [1] A. L. Barabási and R. E. Crandall, “Linked: the new science of
As shown in Figure 6, the results of the experiments networks,” American Journal of Physics, vol. 71, no. 4, pp. 409–
all indicate that the effect of superposition is obvious, and 411, 2003.
the higher the threshold, the fewer the nodes affected. For [2] M. E. J. Newman, “The structure and function of complex
example, when the infection rate is 0.5, the number of networks,” SIAM Review, vol. 45, no. 2, pp. 167–256, 2003.
abandoned nodes with the threshold 1 is almost 4 times [3] A. Barrat, M. Barthlemy, and A. Vespignani, Dynamical pro-
more than that with threshold 3, which even goes to 19 times cesses on complex networks, 2008.
when compared with the result of the threshold 5. The results [4] R. Cohen and S. Havlin, “Complex networks: structure, robust-
accurately explain why information spreads widely in online ness and function. London, UK,” Cambridge University Press,
social networks. With the explosion of Internet and we- 2010.
media, the spreading of information costs little investigation [5] P. Erdős and A. Rényi, “On the evolution of random graphs,”
and individuals spread information with less filters. Such less Transactions of the American Mathematical Society, vol. 286, no.
frequent contact to information leads to the rapid spreading. 1, pp. 257–274, 2011.
Memory superposition factor in our model stands for the [6] D. J. Watts and S. H. Strogatz, Collective Dynamics of ’Small-
number of contacts. World’ Networks, 2011.
[7] A. Barabasi and R. Albert, “Emergence of scaling in random
networks,” Science, vol. 286, no. 5439, pp. 509–512, 1999.
5. Conclusions [8] M. Li, Y. Fan, D. Wang, D. Li, J. Wu, and Z. Di, “Small-world
effect induced by weight randomization on regular networks,”
In summary, we introduce an OSIS model to describe Physics Letters A, vol. 364, p. 488, 2007.
information spreading in online social networks. Firstly, we [9] Q. Jin, L. Wang, C. Xia, and Z. Wang, “Spontaneous symmetry
propose OSIS by combining epidemics SEIR model with indi- breaking in interdependent networked game,” Scientic Reports,
vidual’s cognitive psychology. Secondly, a cellular automata vol. 4, p. 4095, 2015.
model of information transmission is designed, which pro- [10] R. Pastor-Satorras and A. Vespignani, “Epidemic spreading in
vides a computational method for information spreading scale-free networks,” Physical Review Letters, vol. 86, no. 14, pp.
from a microscopic perspective. Additionally, we use small- 3200–3203, 2001.
world network to simulate online social network, which [11] N. Bailey, “The mathematical theory of infectious diseases and
verifies the spreading and evolution of information. The its applications,” The Journal of Immunology, vol. 34, no. 5, pp.
experimental results reveal that our model is consistent with 955-956, 1978.
the propagation of information in online social networks. [12] R. M. Anderson and R. M. May, Infectious Diseases of Humans:
We find that individual’s cognition affects the information Dynamics and Control, Oxford University Press, 1991.
spreading behavior. When the infection rate is low, cognitive [13] N. C. Grassly and C. Fraser, “Mathematical models of infectious
psychology prevents the spread of information. By contrast, disease transmission,” Nature Reviews Microbiology, vol. 6, no.
when the infection rate is sufficiently high, it promotes trans- 6, pp. 477–487, 2008.
mission. We demonstrated that the explosion of online social [14] R. M. May and A. L. Lloyd, “Infection dynamics on scale-free
network scale and the convenience of we-media can greatly networks,” Physical Review E: Statistical, Nonlinear, and Soft
increase the ability of information dissemination. Meanwhile, Matter Physics, vol. 64, no. 6, Article ID 066112, 2001.
Complexity 9

[15] M. H. A. Biswas, L. T. Paiva, and M. D. R. D. Pinho, “A [36] G. J. Martnez, A. Adamatzky, F. Chen, and L. Chua, “On Soliton
SEIR model for control of infectious diseases with constraints,” Collisions between Localizations in Complex Elementary Cel-
Mathematical Biosciences & Engineering, vol. 11, no. 4, p. 761, lular Automata: Rules 54 and 110 and Beyond,” Complex Systems,
2014. vol. 117, 2013.
[16] X. Fu, A. Passarella, D. Quercia, A. Sala, and T. Strufe, “Online [37] H. F. Gagliardi and D. Alves, “Small-world effect in epidemics
Social Networks,” Computer Communications, vol. 73, pp. 163– using cellular automata,” Mathematical Population Studies. An
166, 2016. International Journal of Mathematical Demography, vol. 17, no.
[17] X. Ke, Z. Sai, C. Hao et al., “Measurement and analysis of online 2, pp. 79–90, 2010.
social networks,” Chinese Journal of Computers, pp. 29–42, 2014. [38] X. Jin, J. Li, and L. Zhang, Online Social Networks Based on
[18] Q. Ying, Z. Nan, and Z. Laijun, “Research on the law of weibo Complex Network Theory and Simulation Analysis, Springer
publicity transmission,” Journal of the China Society for Scientic International Publishing, 2015.
and Technical Information, vol. 31, no. 12, pp. 1299–1304, 2012
(Chinese).
[19] weibo, https://weibo.com/.
[20] F. Xiong, Y. Liu, Z. J. Zhang, J. Zhu, and Y. Zhang, “An
information diffusion model based on retweeting mechanism
for online social media,” Physics Letters A, vol. 376, p. 2103, 2012.
[21] D. Centola, “The spread of behavior in an online social network
experiment,” Science, vol. 329, no. 5996, pp. 1194–1197, 2010.
[22] H. Wu, A. Arenas, and S. Gmez, “Influence of trust in the
spreading of information,” Physical Review, vol. 95, 2016.
[23] A. L. Hill, D. G. Rand, M. A. Nowak, and N. A. Christakis,
“Infectious disease modeling of social contagion in networks,”
PLoS Computational Biology, vol. 6, no. 11, Article ID e1000968,
15 pages, 2010.
[24] L. Lü, D.-B. Chen, and T. Zhou, “Small world yields the most
effective information spreading,” New Journal of Physics, p.
825834, 2011.
[25] J. X. Zhang, D. B. Chen, Q. Dong, and Z. D. Zhao, “Identifying
a set of influential spreaders in complex networks,” Scientic
Reports, vol. 6, 2016.
[26] W. Wang, M. Cai, and M. Zheng, “Social contagions on
correlated multiplex networks,” Physica A Statistical Mechanics
& Its Applications, vol. 499, 2017.
[27] W. Wang, X. L. Chen, and L. F. Zhong, “Social contagions with
heterogeneous credibility,” Physica A Statistical Mechanics & Its
Applications, 2018.
[28] W. Wang, M. Tang, S. H. Eugene, and L. A. Braunstein,
“Unification of theoretical approaches for epidemic spreading
on complex networks,” Reports on Progress in Physics, p. 036603,
2017.
[29] W. Wang, M. Tang, H. F. Zhang, H. Gao, Y. Do, and Z. H.
Liu, “Epidemic spreading on complex networks with general
degree and weight distributions,” Physical Review E Statistical
Nonlinear &a Soft Matter Physics, vol. 90, p. 042803, 2014.
[30] X. Zhu, W. Wang, S. Cai, and H. E. Stanley, “Stanley, Dynamics
of social contagions with local trend imitation,” Scientic Reports,
vol. 8, 2018, Scientic Reports.
[31] X. Zhu, W. Wang, S. Cai, and H. E. Stanley, “Optimal imitation
capacity and crossover phenomenon in the dynamics of social
contagions,” Journal of Statistical Mechanics Theory & Experi-
ment, Article ID 063405, p. 063405, 2018.
[32] A. L. Barabsi, R. Albert, and H. Jeong, “Mean-field theory for
scale-free random networks,” Physica A, vol. 272, p. 173, 1999.
[33] S. A. Pandit and R. E. Amritkar, “Characterization and control
of small-world networks,” Physical Review E: Statistical, Nonlin-
ear, and Soft Matter Physics, vol. 60, no. 2, pp. R1119–R1122, 1999.
[34] J. Von Neumann, Papers of John Von Neumann on Computing &
Computer Theory, vol. 1, 1951.
[35] B. Chopard and M. Droz, Cellular Automata for Modeling
Physics, Cambridge University Press, Cambridge, UK, 1998.