Volume 8, Issue 5, May 2023 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Accomodation Recommendation and

Booking for Students
Sakshi Rane1, Shraddha Deskhmukh2, Nidhee Agarwal3, Aditya Verma4, Dr. Amol Dhakne5
Associate Professor 5
Computer Department, D Y Patil Institute of Engineering
Management and Research Pune, India

Abstract:- Students in Various colleges travel from far the request to the backend), which deal with the data. For
away places for a better education. Here they often face backend the best could be NodeJs, and database management
problem while looking for an affordable and safe system we have used Mongo DB, For API we have used
accommodation. Additionally, users often have to check insomnia.
out the places as there is very little information in
amenities , location and user feedback. Not only this but The methodology used here is k-means clustering. It
also often, people are too tired to prepare home-cooked allows us to cluster the data into different groups and a
meals due to the fast-paced and busy environment where convenient way to discover the categories of groups in the
they live. Furthermore, even if you eat only homemade unlabelled dataset on its own without the need for any
food every day, you are still likely to want to go out to eat training. It is a clustering algorithm that helps in forming
for social or recreational purposes from time to time. clusters and classification of data in various groups.
Despite this, it's common knowledge that food is an
important aspect of anyone's lifestyle, no matter where II. EASE OF USE
they live. Assume, for example, someone has just moved This is the software configuration in which the project
to a new home. Their preferences and tastes are already was shaped. The programming language used, tools used, etc
established. If a student lives nearby her favourite outlet, are described here.
this will save her a lot of trouble and help her save money.  Python
Students nowadays also have to take care of their health
 REACT JS
and fitness. Hence, recommendation for affordable gyms
has also become a requirement. Based on incoming  Node JS
students' preferences for facilities, budget, and proximity  Database management system.
to the location, this project uses K-Means Clustering to  For data - numpy and pandas package. 15 • For plotting -
find the most suitable accommodation, restaurant as well matplotlib package & seaborn packages
as gyms for them in Akurdi (Pune).  For geospatial - geopy, folium.
 •For machine learning – sklearn, scipy
Keywords:- Geolocation data analysis, k- means clustering, python,
recommendation. A. back-end
This is for the working of backend development which is
I. INTRODUCTION configuration, administration and management of databases
and servers
The majority of Indian migrants migrate for study
 NodeJs- Create a server side web application
purposes, with 2.9 million moving within the state from their
previous residence for less than 5 years. There are around  MongoDB- Used as a backend data store
50,000 registered hostels and accommodation services  Insomnia- Open source application For the api
available in India. connection, designing, debugging and testing
 REST
Geolocational data analysis is the process for analysing
data based geographic location. It is possible to make III. METHODOLOGIES
recommendations for accommodations based on the location
of the user and the location of available accommodations  K-Means Clustering
through the use of geolocational data analysis. Using this A. Introduction
type of analysis it can also be used to identify various trends K-Means Clustering is an unsupervised learning
and patterns in the behaviour and preferences of the user. technique that groups unlabelled data into distinct clusters.
Overall, geolocation data analysis can significantly improve Here K defines the number of groups that should be created
the accuracy rate of our proposed system. in this process, if K=2 there will be 2 groups, if K=3 there
will be 3 groups and so on. It allows us to cluster the data
In our proposed system we are using technologies that
provide maximum benefit and usability regarding into different groups and a convenient way to discover the
categories of groups in the unlabelled dataset on its own
geolocational data analysis. For front-end we are making use
without the need for any training. The main objective of the
of REACT, it is a front-end JavaScript library. We are using
K-Means algorithm is to minimize the sum of distances
it to make our website’s interface reactive and more
between the points and their respective cluster centroid. This
userfriendly. React is capable of making API calls (sending
algorithm is based on centroids. Each cluster is assigned a

IJISRT23MAY1418 www.ijisrt.com 3804

Volume 8, Issue 5, May 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
centroid and tries to minimize the distance between each
point to cluster. Here our data points are the locations of
particular place i.e. the accommodation, hostel, hotel,
restaurant, mess, gyms, library etc. and the clusters so
formed are according to the filters applied by the user
according to their preferences and choices.

B. Steps in k-means clustering  Repeat steps 3 and 4

We then repeat steps 3 and 4:
 Choose the number of clusters k
The first step in k-means is to pick the number of clusters,
k. 2. Select k random points from the data as centroids

Next, we randomly select the centroid for each cluster.

Here, the red and green circles represent the centroid for
these clusters.
 Stopping criteria for k-means clustering
There are essentially three stopping criteria that can be
adopted to stop the K-means algorithm:
 Centroids of newly formed clusters do not change
 Points remain in the same cluster
 Maximum number of iterations is reached
 Assign each point to the nearest cluster centroid.
Every point is assigned to the nearest cluster centroid. IV. IMPLEMENTATION
The points closest to the red point are assigned to the red
cluster, while the ones closest to the green point are assigned K-means clustering or any other algorithm in machine
to the green cluster. learning can be implemented by either using Jupyter
Notebook or Google Collaboratory or Colab which provides
us with necessary interface and runtime. These are
opensource web-based interactive computing platforms;
Jupyter Notebook is an incredibly powerful tool for
interactively developing and presenting data science
projects. Colab allows anybody to write and execute
arbitrary python code through the browser, and is especially
well suited to machine learning, data analysis and education.
 Calculate the centroids of freshly generated clusters
again. A. BASIC PROCEDURE
In the next step, we will compute the centroids of the Implementing k-means clustering algorithm in python
newly created clusters after all of the points have been needs a few libraries to be installed first.
allocated to either cluster. Centroids have been revised as
indicated by the red and green crosses.  So the first step is to install the necessary libraries
needed
To be able to implement k-means we have libraries like
sklearn, cluster, numpy, pandas, etc.

Fig. 1: Installing libraries

IJISRT23MAY1418 www.ijisrt.com 3805

Volume 8, Issue 5, May 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
 Pre-processing of data
 Select the dataset to be used Once we have the desired the data, it still contains some
We find the appropriate dataset that is to be used for unwanted information, we remove this by pre-processing
evaluation and finding answers. We then upload this dataset and cleaning the data.
and begin with data exploration.
 Data visualization
We use various plotting techniques to visualize our data
before applying the algorithm to get an overview of the
dataset.

Fig. 2: Data Visualization

 Geolocational data
We then using the API collect and feed the geolocational inputs to the model and expect an output.

Fig. 3: Geolocational data

B. Application Programming Interface(API) mainly apo’s are inherited from the Foursquare, which is
mostly used for building the review app structure or ride
 INTRODUCTION sharing/navigation to the user. Third-party apps can connect
In many places, geolocation services are becoming to Foursquare and augment the official app's functionality.
essential for a variety of applications and disciplines. When Developers can gain access to and collect useful data for
the consumer is Many websites that are frequently visited use analysis. The input given to this model is the current location
geolocation data for targeted localized advertising, localized and the radius of the search for which you need the
content (such as local news and weather), and adherence with recommended places. And the output gives all the popular
local laws occasionally without the user being aware of it. places around a given location and python to visualize this
APIs play a key part in enabling these functions. In the stuff.
context of API’s, the word Application refers to any software
with a distinct function. Interface can be thought of as a  STEPS IN API IMPLEMENTATION USING
contract of service between two applications. This defines FOURSQUARE API
how the two communicate with each other using requests and  Data Requirement: Visit the foursquare developer
responses. API automation is used as API’s are the internal website and a developer account needs to be created
connections which are present on both web and mobile Client id and Client secret credentials will be provided
application, which is hidden from the user, but critical for the .Now to explore a location we use this URL where
application to perform functions. It is analogous to human latitude and longitude are current coordinates which we
body containing blood vessels and nervous system. Hence can get using geopy library. Limit is the number of
when performing automation in this layer it requires specific results which will be fetched.
targeted information, inputs and outputs. API- in this project

IJISRT23MAY1418 www.ijisrt.com 3806

Volume 8, Issue 5, May 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig. 4: URL of coordinates

 Model Development: 1)Importing all essential python libraries 2)Defining the foursquare API credentials. 3)Read the current
location from the input and converting it to Coordinates.

Fig. 5: Model Development

 Fetching data from the foursquare API, returns output as JSON data.

Fig. 6:

 Extracting the relevant data from Json which fetches the V. PROPSED SYSTEM
raw data.
 Cleaning the data hence converting it into data frames, so The proposed system fosters hostels, apartments, and
its ready and visualization takes place houses, and it additionally gives data on those houses, flats,
and hostels (for example, images and contact information).

It suggests budget-friendly lodging. It has a lot of

houses on our budget. It also has the potential in the future

IJISRT23MAY1418 www.ijisrt.com 3807

Volume 8, Issue 5, May 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
to propose restaurants, gyms, and other services depending downside when two circular clusters with the same mean
on the user's budget and tastes. It delivers accurate have different radii. K-Means defines a cluster's median
recommendations with no gaps. In this project, we are using median values and does not distinguish between the
employing the K-means technique, although it has a two clusters. When the sets are noncircular, it likewise fails.

Fig. 7: UML data

VI. CONCLUSION [3.] M. N. Y. Utomo, T. B. Adji and I. Ardiyanto,

"Geolocation prediction in social media data using
The similar existing models provide a few facilities that text analysis: A review," 2018 International
are not sufficient, some newly added features in our model Conference on Information and Communications
helps students in not only finding the best accommodation Technology (ICOIACT), Yogyakarta, Indonesia,
but also other places like mess, restaurants, gyms, library etc. 2018, pp. 84-89, doi:
The use of k-means clustering helps in finding the 10.1109/ICOIACT.2018.8350674.
appropriate location according to filters and choices, this [4.] S. R. Kishan, K. S. Kakunuri, A. Raj Parasa, V.
technique can cluster the migrant data and geolocational Patlolla and P. Gamini, "Exploratory Analysis on Geo-
data. Our website represents the analysed model in form of a Locational Data," 2022 International Conference on
plotted-map. This website will prove to be best suited for Automation, Computing and Renewable Systems
students to ease their shifting process to a new city with the (ICACRS), Pudukkottai, India, 2022, pp. 1-4,
recommendations filtered according to their interests. Future doi: 10.1109/ICACRS55517.2022.10029164.
research will focus on taking multiple locations but not a [5.] M. N. Y. Utomo, T. B. Adji and I. Ardiyanto,
fixed location and applying clustering techniques on it. "Geolocation prediction in social media data using
text analysis: A review," 2018 International
ACKNOWLEDGMENT Conference on Information and Communications
This paper was partially supported by Dr. D Y Patil Technology (ICOIACT), Yogyakarta, Indonesia,
Institute of Engineering Management and Research, Akurdi. 2018, pp. 84-89, doi:
This paper was created under the guidance of Dr. Amol 10.1109/ICOIACT.2018.8350674.
Dhakne. [6.] Singh R, Manchanda N, Mishra R. Internal student
migration in India:
REFERENCES [7.] Impact of the COVID-19 crisis. Asian Pac Migr J.
2022 Dec;31(4):454–77. doi:
[1.] S. -H. Wu, J. Hsiao, Y. -S. Wu and J. -T. Jeng, "AI 10.1177/01171968231154590. Epub 2023 Feb
Food Recommendation Systems," 2022 IET [8.] PMCID: PMC9944466.
International Conference on Engineering
Technologies and Applications (IET-ICETA),
Changhua, Taiwan, 2022, pp. 1-2, doi: 10.1109/IET-
ICETA56553.2022.9971598.
[2.] P. T. Chiou and G. S. Young, "Implementing
Recommendations of Accessibility Technology
Guidelines — The Quantitative Effects and Benefits
it Offers to Non-disabled Students," 2017
International Conference on Computational Science
and Computational Intelligence (CSCI), Las Vegas,
NV, USA, 2017, pp. 1137-1142, doi:
10.1109/CSCI.2017.198.

IJISRT23MAY1418 www.ijisrt.com 3808