Design of a Novel Machine Learning Algorithm to Predict Number of Book Copies Required in Library - A Review

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 02 | Feb 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 88
Design of a Novel Machine Learning Algorithm to Predict Number of Book Copies
Required in Library - A Review
Darshan D M1, Pavan N S2, Pramod Bhat3, Sagar S M4, Prof. Sunanda5
1,2,3,4,5Department of Computer Science and Engineering, DSCE, Bengaluru
---------------------------------------------------------------------***----------------------------------------------------------------------
Abstract: "Libraries store the energy that fuels the
imagination. They open up windows to the world and
inspire us to explore and achieve, and contribute to
improving our quality of life." [1.1]
Libraries are the storage of knowledge. The Library
Management system is a very useful and needy tool in
colleges and institutions. Present Library systems are
managed using book-log entry and manual segregation of
the books. The library admin has to monitor the book
issue and manually work on collecting back the issued
books with applicable late fees. With the increase in the
issue of books, it is also to search for any log entries. Some
institutions and libraries use computerized log-entry. But
still, they fail to provide online book search to students
and have multi-user interactions like the interaction of
students, teachers, library admin in a single platform. With
the advancement in technology and everything at the tip of
the hand, following the old custom of book-log entry will
not be efficient for maintenance of today’s Libraries with
such a huge collection of books and data. A platform is also
required to organize E-books and make them available to
the readers. The project aims in designing a multi-user
login based web-application with the integration of
Machine Learning models to fasten and make library usage
handy. This paper is on a literature survey that is carried
out to learn and analyze the area of our research. The
survey includes studying the various concepts that are
involved in the “Design of a Novel Machine Algorithm to
predict the number of book copies required in Library”.
The study covers the analysis and study of various kinds of
prediction algorithms associated with book sales or
rentals, recommendation algorithms, multi-user website
designing with security, optimizing web page performance.
Various algorithms have been studied in the research
survey including Artificial neural networks, random
forests, etc. Various Novel algorithms were also found
during the research.
Keywords: Library management; Web Application;
Prediction model; recommendation model; Multi-User
Login
Introduction
The literature survey was carried out taking into
considerations the key requirements of the project which
include Web Application with different user credentials. AI
model for recommendation system and predictive analysis.
Literature survey also aided in one of the major challenge
that is payment gateway integration.
The literature survey was carried out taking into
considerations the key requirements of the project which
include Web Application with different user credentials. AI
model for recommendation system and predictive analysis.
Literature survey also aided in one of the major challenge
that is payment gateway integration.
Paper [1] aided us in studying the algorithm and method
that can be implemented to predict the number book
copies that would be required based on the frequency of
the book issue.
Paper [2] referred above aided in understanding that the
normal recommendation system that is used in shopping
or any other cases can not directly be used in book
recommendation system. Instead a tagging and
recommend method must be developed for book
recommendation system.
Paper [3] helped us in planning the sub-systems and the
modules for implementing the project. The paper is
equipped with use case diagrams and E-R diagram which
aided us in designing of our use cases that are shown in
this document.
Apart from the above main papers, many research and
journal papers have aided in framing the project idea
which are discussed in Literature Survey.
Literature Survey
Following are the papers which were referenced to carry
out the literature survey. Following is a brief idea on the
author, concept of the paper, model used and also its
advantages and disadvantages.
The paper presents the concept of prediction analysis of
rental books data. A Novel model is designed called DPA-
LRPD Model(Data and Predictive Analysis based on
Library rental Book Data)[4] and trained. Advantages: It
has discovered hidden patterns and knowledge in the data
set that aided in predicting future demand.
Disadvantages: The paper limits and analyses less data
and could be improved to gain more knowledge like
sentiment and semantic analysis on rental user reviews
data. [4]
The paper by Satyendra Kumar Sharma, et al (2019),
analyses different algorithms to predict the book sales at
Amazon Market Place. Multiple algorithms including
Artificial Neural Network algorithm, Decision Tree
algorithm and random forest algorithm have been

analyzed. Advantages: Review sentiments doesn’t have
very great impact unless the reviews are more in number.
Disadvantages: The Study limits with very less predictor
variables. Genre, characteristics of book, country can also
be considered. [5]
The paper by Takanori Kuroiwa et al(2007), is a research
on Dynamic book recommendations using web services
and virtual library enhancement which is carried out using
Collaborative filtering,Content based methods and XML
database. Advantages: Using this model we can increase
the purchase of book and recommend the desired book to
user. Disadvantages: The mentioned algorithm must be
further investigated for better ease for recommendation
system.[10]
The paper by Nursulthan Kurmashov et al(2012) , is a
research on Online book recommendation System which is
carried out by using Collaborative filtering and Mysql
database . Advantages: using this algorithm we can
recommend books based on users ratings, interest to
increase the purchase of book. Disadvantages:
Mentioned model can recommend books based on only
user ratings this cant be implemented in all the cases it
varies from user to user.[11]
This paper by Muzaffer Ege Alper et al(2012) , is a
research on Personalized recommendation using joint
probabilistic model of users. It is done by using a
probabilistic personalized model and a latent interest
model. Advantages: It indicates the interest context of an
user which in turn help in designing a personalized
recommendation system. Disadvantages: As the
recommendation varies from user to user we cant predict
it properly using this mentioned models so we must
implement the other models or enhance the existing
model for better ease of the results for better prediction.[16]
The paper by Xiantao Jiao et al(2012) , is a research on A
semanamtic tagging system for biomedical Articles which
is carried out by using PubMed tagging model.
Advantages: This Optimised PubMed tagging system has
been designed with semantics web technologies, it is more
efficient than the system which has complex semantics
elements[17] Disadvantages: The two kind of TBox and
ABox semantics has to be separated to further improve the
system performance. Lacks with very less strategies used
to implement better quality control of semantic data with
large semantic repository.[17]
This paper by Ipek Tatli Aysenur Birturk(2011) , is a
research on a tag based hybrid music recommendation
which uses relations and multi domain functions for
analysis , Latent Symantic Analysis(LSA). Advantages :
With restpect to LSA it is used to order the words and the
morphology. LSA is used for large document
categorization and text summarization. Disadvantages:
Creating a dataset and ontology takes more time. Tags that
are less in number can be neglected.[18]
This paper by Kensuke Baba et al(2016) , is a research on
predicting the use of books in university library by using
synchronous obsolescence . Advantages: It was able
predict the number as accurately as a diachronic
obsolescence algorithm.It is used to predict the loans
number accurately in case of incomplete circulation of
data[19]. Disadvantages: It does not consider differences
number loans change among the subject fields which may
cause significant difference in future use.[19]
The paper by Craig Silverstein et al(2014) , is a research
on predicting the book used for physical storage using a
decision tree. Advantages: The research could explore
several approaches for the building of decision tree to
know which approach is better and good. Disadvantages:
The mentioned algorithm/model is more complicated
which is not implemented for large values found in library
data set and they slow downs the results.[20]
This paper by Lina Zhou et al(2017) , is a research on the
opportunities and challenges that are faced during
machine learning on big data. Advantages: Many
parameters have been studied find the opportunities in
machine learning. Disadvantage: There are open research
issues like cleaning and compressing of the big data. Only a
small scale distributed feature selsction is done.[22]
The paper by Shih-Ting Yang et al(2012), is on a model for
analysis of book inquiry and and a book recommendation
of libraries. It uses the models and algorithms like book-
acquisition recommendation model[25], Advantages:
Using this algorithms a relationship between book
categories and the keywords can be developed. It also
helps in developing a web based book recommendation[25]

Table Representation of Literature Survey
Ref
No
Authors Concept/Idea of
Research
Algorithm/ M
odel
Advantages Disadvantages
[24] Imran
et al(2020)
An effective
planning of waste
management
using predictive
data analysis.
The model is
evaluated with
Mean Absolute
Error
checking(MAE)
and also with
root mean
square and
mean absolute
percentage
error.
It helps to produce
waste information
very early that Is
utilized by the waste
management
authorities for
effective designing.
The research limits
with online
providing data but
can be stretched to
do many
applications.
[26] Noor Ifada et
al(2019)
A Library
recommendation
system.
Probabilistic
keyword model,
collaborative
filtering.
It builds a model
with book
circulation records
and keyword data. It
employs a
probabilistic
technique.
The paper could have
also worked on
attribute model with
different attributes
combination and
could solve the
sparsity problem.
[27] Liu xin et al
(2013)
Book
recommendation
model based on
the borrowing
records of the
users.
Collaborative
filtering(CF)
algorithm
Focus on how to
implement efficient
and
accurate CF
algorithms for the
academic library. By
transforming the
rating information
from the borrowing
records, we can use
the traditional CF
algorithms to predict
how long the reader
will borrow the book
It lacks on on how to
evaluate SR under
the individual CF
algorithms and the
blending methods as
while we use NN
with KNNuser and
LFM, SR is just
2.95%. efficient and
more blending
methods is not
studied.
Findings
A) Artificial Neural Network algorithm is best for the
prediction of number of book copies required in the
library.[1]
B) A tagging-based algorithm must be used for book
recommendation system.[2]
C) Dynamic book recommendations using web services
and virtual library enhancement [10] aides in increasing
the book issue and sales.
D) Studied the framework and the ability to model the
users jointly. This would help in designing the personal
recommendation system[16]
Conclusions
Based on the survey carried out, many ideas have been
brought up which are pointed out in the findings section.
Some of the papers have designed concepts and models
which can be directly implemented and some others which
can be altered according to the problem. The project
implementation will be carried out based on the findings
and outcome of the research.
The outcome Library Management System project idea is
an online Web Application based solution to maintain

majority of the library activities like book search, book
issue and return management. The system also has
additional features like separate login modules for admin,
teachers and students. It also has a ML model for book
recommendation to students and to predict and analyse
the number of book copies required based on the issue
frequency.
The application is aided with the real time data set from
College library which will ease the data collection. The
model is trained and deployed onto the Web application
which is designed with React and Javascript. The complete
system will be developed and the prototype will be
deployed on a local host.
References
[1] Dersin Daimri, Mwnthai Narzary, N Mazumdar, Amitav
Nag, (2021). ‘A Machine Learning Based Book Availability
Prediction Model for Library Management System’,
Library Philosophy and Practice (e-journal). 4982
[1.1]https://www.librarianshipstudies.com/2018/05/quo
tes-libraries-librarians-library-information-science.html
[2] Punit Gupta, Ravi Shankar Jha, (2015). ‘Tagging Based
Evolving Recommendation System for Digital Library
System’, 4th International Symposium on Emerging
Trends and Technologies in Libraries and Information
Services, 978-1-4799-7999-9/15/$31.00 ©2015 IEEE
[3] Md Robiul Alam Rabel, Subrato Bharati, Prajay Podder,
M. Rubaiyat Hossain Mondal, ‘Integrated Cloud Based
Library Management in Intelligent IoT driven
Applications’. (2021). Fog, Edge, and Pervasive Computing
in Intelligent IoT Driven Applications, First Edition.
[4] N.Iqbal, Faisal Jamil, Shabir Ahmad, Dohyeun Kim,
‘Toward Effective Planning and Management using
predictive analytics based on rental book data of academic
libraries’. IEEE, DOI: 10.1109/ACCESS.2020.2990765
[5] Satyendra Kumar Sharma, Swapnajit Chakraborti,
Tanaya Jha, ‘Analysis of book sales prediction at Amazon
marketplace in India: a machine learning approach’(2019),
Indormation Systems and e-Business Management, DOI:
10.1007/s10257-019-00438-3
[6] Wenyu Li, D.Chen, Xiaoyu Duan, Changchang Huang,
Yayun Lu, Xuemei Hu, ‘The design of disciplinary book
recommendation system based on android: a view of
extra-curricular activities’(2019), International
Conference on Communications, Information System and
Computer Engineering(CISCE), IEEE, DOI
10.1109/CISCE.2019.00038
[7] Brusilovsky, Peter; Kobsa, Alfred; Nejdl, Wolfgang
(2007). ’User Profiles for Personalized Information
Access’. Lecture Notes in Computer Science, The Adaptive
Web Volume 4321 || DOI 10.1007/978-3-540-72079-9_2
[8] eCommerce Payment Gateway Integration for
Your Website & App (with Video-Guides), Zoolatech Team,
https://zoolatech.com/blog/ecommerce-payment-
gateway-integration-for-your-website-app-with-video-
guides/
[9] Yongcheng Luo, Jiajin Le, Huilan Chen, ‘A Privacy-
Preserving Book Recommendation Model Based on Multi-
Agent’, (2009), Second International Workshop on
Computer Science and Engineering, IEEE, DOI
10.1109/WCSE.2009.200
[10] Takanori Kuroiwa and Subhash Bhalla, ‘Dynamic
Personalization for Book Recommendation System using
Web Services and Virtual Library Enhancements’, Seventh
Intenational Conference on Computer and Information
Technology, DOI 10.1109/CIT.2007.72
[11] How to Integrate Machine Learning into Web
Applications with Flask, Analytics Vidhya,
https://www.analyticsvidhya.com/blog/2020/09/integra
ting-machine-learning-into-web-applications-with-flask/
[12] Deploying Deep Learning Models Part 1:
Preparing the Model, PaperspaceBlog,
https://www.analyticsvidhya.com/blog/2020/09/integra
ting-machine-learning-into-web-applications-with-flask/
[13] Ayaz Ahmad Sofi, Atul Garg, ‘Analysis of various
techniques for improving Web performance’, (2015),
International Journal of Advance Research in Computer
Science and Management Studies, Vol 3.
[14] Jatinder Manhas, ‘A study of factors affecting
websites page loading speed for efficient Web
performance’,(2013) Internation Journal of Computer
Science and Engineering
[15] N.Kurmashov, K. Latuta, Abhay Nussipbekov,
‘Online Book Recommendation System’.
[16] Muzaffer Ege Alper, Sule Gunduz Oguducu,
‘Personalised Recommendation in Folksonomies using a
Joint Probabilistic Model of Users, Resources and Tags.’
(2012)
[17] Xiantao Jiao, Yue Chen, ‘A Semantic Tagging
System for Biomedical Articles’, (2010), International
Conference of Biomedical Engineering and Informatics.
[18] Ipek Tath, Aysenur Birturk, ‘A Tag-bsed Hybrid
Music Recommendation System Using Semantic Relations
and Multi-domain Information’, (2011), 11th IEEE
International Conference on Data Mining Workshops.
[19] Kensuke Baba, Toshiro Minami, Tetsuya Nakatoh,
‘Predicting Book Use in University Libraries by
Synchronous Obsolescence’, (2016) 20th International
Conference on Knowledge Based and Intelligent
Information and Engineering Systems.

[20] Craig Silverstein, Stuart M Shieber, ‘Predicting
Individual Book Use for Off-site Storage using Decision
Trees’ (2014).
[21] G.Edward Evans, ‘Book selection and book
collection usage in academic libraries’, (1970), The Library
Quaterly, Vol.40, No.3.
[22] Lina Xhou, Shimei Pan, jianwu Wang, ‘Machine
Learning on Big Data: Opportunities and Challenges’
[23] Imran, Shabir Ahmad, Do Hyeun Kim, ‘Quantum
GIS based descriptive and Predictive Data Analysis for
Effective Planning of Waste Management’(2020)
[24] Chia-Nan Ko, Cheng-Ming Lee, ‘Short-term load
forecasting using SVR(Support vector regression)- based
radial basis function neural network with dual extended
Kalman filter’(2012)
[25] Shis-Tang Yang, Ming-Chien Hung, ‘A model for
book inquiry history analysis and book-aquisition
recommendation of libraries’, (2012)
[26] Noor Ifada, et al.., ‘Enhancing the Performance of
Library Book Recommendation System by Employing the
Probabilistic-Keyword Model on a Collaborative Filtering
Approach’(2019), 4th International Conference on
Computer Science and Computational Intelligence.
[27] Liu Xin, E Haihong, et al.., ‘Collaborative Book
Recommendation Based on Readers’ Borrowing
Records’(2013), International Conference on Advanced
Cloud and Big Data.
[28] Dwi H. Widyantoro, Thomas R. loerger, John Yen,
‘Learning User Interest Dynamics with a three-descriptor
Representation’(2001), Journal of the American Society for
Information Science and Technology.
[29] Barry Crabtree, Stuart J Soltysiak, ‘Identifying and
tracking changing interests’(1998), Intelligent Systems
Research Group.

Design of a Novel Machine Learning Algorithm to Predict Number of Book Copies Required in Library - A Review

More Related Content

Similar to Design of a Novel Machine Learning Algorithm to Predict Number of Book Copies Required in Library - A Review (20)

More from IRJET Journal (20)

Recently uploaded (20)

Design of a Novel Machine Learning Algorithm to Predict Number of Book Copies Required in Library - A Review