Experience Seamless Full Ebook Downloads for Every Genre at textbookfull.

com

Computational Methods and Data Engineering:

Proceedings of ICMDE 2020, Volume 1 Vijendra Singh

https://textbookfull.com/product/computational-methods-and-
data-engineering-proceedings-of-
icmde-2020-volume-1-vijendra-singh/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Computational Methods and Data Engineering: Proceedings of

ICMDE 2020, Volume 2 Vijendra Singh

https://textbookfull.com/product/computational-methods-and-data-
engineering-proceedings-of-icmde-2020-volume-2-vijendra-singh/

textboxfull.com

Software Engineering Perspectives in Intelligent Systems:

Proceedings of 4th Computational Methods in Systems and
Software 2020, Vol.1 Radek Silhavy
https://textbookfull.com/product/software-engineering-perspectives-in-
intelligent-systems-proceedings-of-4th-computational-methods-in-
systems-and-software-2020-vol-1-radek-silhavy/
textboxfull.com

Data Management, Analytics and Innovation: Proceedings of

ICDMAI 2020, Volume 1 Neha Sharma

https://textbookfull.com/product/data-management-analytics-and-
innovation-proceedings-of-icdmai-2020-volume-1-neha-sharma/

textboxfull.com

CFA 2020 Level 1 Volume 1 Ethics and Quantitative Methods

Various

https://textbookfull.com/product/cfa-2020-level-1-volume-1-ethics-and-
quantitative-methods-various/

textboxfull.com
Software Engineering Perspectives in Intelligent Systems:
Proceedings of 4th Computational Methods in Systems and
Software 2020, Vol.2 Radek Silhavy
https://textbookfull.com/product/software-engineering-perspectives-in-
intelligent-systems-proceedings-of-4th-computational-methods-in-
systems-and-software-2020-vol-2-radek-silhavy/
textboxfull.com

Coulson and Richardson’s Chemical Engineering, Fourth

Edition: Volume 3A: Chemical and Biochemical Reactors and
Reaction Engineering R. Ravi
https://textbookfull.com/product/coulson-and-richardsons-chemical-
engineering-fourth-edition-volume-3a-chemical-and-biochemical-
reactors-and-reaction-engineering-r-ravi/
textboxfull.com

Data Engineering and Applications Volume 1 Rajesh Kumar

Shukla

https://textbookfull.com/product/data-engineering-and-applications-
volume-1-rajesh-kumar-shukla/

textboxfull.com

Proceedings of International Conference on Computational

Intelligence and Data Engineering Proceedings of ICCIDE
2018 Nabendu Chaki
https://textbookfull.com/product/proceedings-of-international-
conference-on-computational-intelligence-and-data-engineering-
proceedings-of-iccide-2018-nabendu-chaki/
textboxfull.com

Advances in Computational and Bio-Engineering: Proceeding

of the International Conference on Computational and Bio
Engineering, 2019, Volume 1 S. Jyothi
https://textbookfull.com/product/advances-in-computational-and-bio-
engineering-proceeding-of-the-international-conference-on-
computational-and-bio-engineering-2019-volume-1-s-jyothi/
textboxfull.com
Advances in Intelligent Systems and Computing 1227

Vijendra Singh
Vijayan K. Asari
Sanjay Kumar
R. B. Patel Editors

Computational
Methods
and Data
Engineering
Proceedings of ICMDE 2020, Volume 1
Advances in Intelligent Systems and Computing

Volume 1227

Series Editor
Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences,
Warsaw, Poland

Advisory Editors
Nikhil R. Pal, Indian Statistical Institute, Kolkata, India
Rafael Bello Perez, Faculty of Mathematics, Physics and Computing,
Universidad Central de Las Villas, Santa Clara, Cuba
Emilio S. Corchado, University of Salamanca, Salamanca, Spain
Hani Hagras, School of Computer Science and Electronic Engineering,
University of Essex, Colchester, UK
László T. Kóczy, Department of Automation, Széchenyi István University,
Gyor, Hungary
Vladik Kreinovich, Department of Computer Science, University of Texas
at El Paso, El Paso, TX, USA
Chin-Teng Lin, Department of Electrical Engineering, National Chiao
Tung University, Hsinchu, Taiwan
Jie Lu, Faculty of Engineering and Information Technology,
University of Technology Sydney, Sydney, NSW, Australia
Patricia Melin, Graduate Program of Computer Science, Tijuana Institute
of Technology, Tijuana, Mexico
Nadia Nedjah, Department of Electronics Engineering, University of Rio de Janeiro,
Rio de Janeiro, Brazil
Ngoc Thanh Nguyen , Faculty of Computer Science and Management,
Wrocław University of Technology, Wrocław, Poland
Jun Wang, Department of Mechanical and Automation Engineering,
The Chinese University of Hong Kong, Shatin, Hong Kong
The series “Advances in Intelligent Systems and Computing” contains publications
on theory, applications, and design methods of Intelligent Systems and Intelligent
Computing. Virtually all disciplines such as engineering, natural sciences, computer
and information science, ICT, economics, business, e-commerce, environment,
healthcare, life science are covered. The list of topics spans all the areas of modern
intelligent systems and computing such as: computational intelligence, soft comput-
ing including neural networks, fuzzy systems, evolutionary computing and the fusion
of these paradigms, social intelligence, ambient intelligence, computational neuro-
science, artificial life, virtual worlds and society, cognitive science and systems,
Perception and Vision, DNA and immune based systems, self-organizing and
adaptive systems, e-Learning and teaching, human-centered and human-centric
computing, recommender systems, intelligent control, robotics and mechatronics
including human-machine teaming, knowledge-based paradigms, learning para-
digms, machine ethics, intelligent data analysis, knowledge management, intelligent
agents, intelligent decision making and support, intelligent network security, trust
management, interactive entertainment, Web intelligence and multimedia.
The publications within “Advances in Intelligent Systems and Computing” are
primarily proceedings of important conferences, symposia and congresses. They
cover significant recent developments in the field, both of a foundational and
applicable character. An important characteristic feature of the series is the short
publication time and world-wide distribution. This permits a rapid and broad
dissemination of research results.
** Indexing: The books of this series are submitted to ISI Proceedings,
EI-Compendex, DBLP, SCOPUS, Google Scholar and Springerlink **

More information about this series at http://www.springer.com/series/11156

Vijendra Singh Vijayan K. Asari
• •

Sanjay Kumar R. B. Patel

•

Editors

Computational Methods
and Data Engineering
Proceedings of ICMDE 2020, Volume 1

123
Editors
Vijendra Singh Vijayan K. Asari
School of Computer Science Department of Electrical and Computer
University of Petroleum and Energy Studies Engineering
Dehradun, Uttarakhand, India University of Dayton
Dayton, OH, USA
Sanjay Kumar
Department of Computer Science R. B. Patel
and Engineering Department of Computer Science
SRM University Delhi-NCR and Engineering
Sonepat, Haryana, India Chandigarh College of Engineering
and Technology
Chandigarh, Punjab, India

ISSN 2194-5357 ISSN 2194-5365 (electronic)

Advances in Intelligent Systems and Computing
ISBN 978-981-15-6875-6 ISBN 978-981-15-6876-3 (eBook)
https://doi.org/10.1007/978-981-15-6876-3
© Springer Nature Singapore Pte Ltd. 2021
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface

We are pleased to present Springer Book entitled Computational Methods and Data
Engineering, which consists of the Proceedings of the International Conference on
Computational Methods and Data Engineering (ICMDE 2020), Volume 1 papers.
The main aim of the International Conference on Computational Methods and
Data Engineering (ICMDE 2020) was to provide a platform for researchers and
academia in the area of computational methods and data engineering to exchange
research ideas, results and collaborate together. The conference was held at the SRM
University, Sonepat, Haryana, Delhi-NCR, India, from January 30 to 31, 2020.
All the 49 published chapters in the Computational Methods and Data Engineer-
ing book have been peer reviewed by three reviewers drawn from the scientific
committee, external reviewers and editorial board depending on the subject matter
of the chapter. After the rigorous peer-review process, the submitted papers were
selected based on originality, significance and clarity and published as chapters.
We would like to express our gratitude to the management, faculty members and
other staff of the SRM University, Sonepat, for their kind support during the
organization of this event. We would like to thank all the authors, presenters and
delegates for their valuable contribution in making this an extraordinary event.
We would like to acknowledge all the members of honorary advisory chairs,
international/national advisory committee members, general chairs, program chairs,
organization committee members, keynote speakers, the members of the technical
committees and reviewers for their work.
Finally, we thank series editors, Advances in Intelligent Systems and
Computing, Aninda Bose and Radhakrishnan for their high support and help.

Editors
Dehradun, India Vijendra Singh
Dayton, USA Vijayan K. Asari
Sonepat, India Sanjay Kumar
Chandigarh, India R. B. Patel

v
Contents

Content Recommendation Based on Topic Modeling . . . . . . . . . . . . . . . 1

Sachin Papneja, Kapil Sharma, and Nitesh Khilwani
Hybrid ANFIS-GA and ANFIS-PSO Based Models for Prediction
of Type 2 Diabetes Mellitus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Ratna Patil, Sharvari Tamane, and Nirmal Rawandale
Social Network Analysis of YouTube: A Case Study on Content
Diversity and Genre Recommendation . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Shubham Garg, Saurabh, and Manvi Breja
Feature Extraction Technique for Vision-Based Indian Sign Language
Recognition System: A Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Akansha Tyagi and Sandhya Bansal
Feature-Based Supervised Classifier to Detect Rumor
in Social Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Anamika Joshi and D. S. Bhilare
K-harmonic Mean-Based Approach for Testing
the Aspect-Oriented Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Richa Vats and Arvind Kumar
An Overview of Use of Artificial Neural Network in Sustainable
Transport System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Mohit Nandal, Navdeep Mor, and Hemant Sood
Different Techniques of Image Inpainting . . . . . . . . . . . . . . . . . . . . . . . 93
Megha Gupta and R. Rama Kishore
Web-Based Classification for Safer Browsing . . . . . . . . . . . . . . . . . . . . . 105
Manika Bhardwaj, Shivani Goel, and Pankaj Sharma

vii
viii Contents

A Review on Cyber Security in Metering Infrastructure

of Smart Grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Anita Philips, J. Jayakumar, and M. Lydia
On Roman Domination of Graphs Using a Genetic Algorithm . . . . . . . 133
Aditi Khandelwal, Kamal Srivastava, and Gur Saran
General Variable Neighborhood Search for the Minimum Stretch
Spanning Tree Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Yogita Singh Kardam and Kamal Srivastava
Tabu-Embedded Simulated Annealing Algorithm for Profile
Minimization Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Yogita Singh Kardam and Kamal Srivastava
Deep Learning-Based Asset Prognostics . . . . . . . . . . . . . . . . . . . . . . . . . 181
Soham Mehta, Anurag Singh Rajput, and Yugalkishore Mohata
Evaluation of Two Feature Extraction Techniques for Age-Invariant
Face Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Ashutosh Dhamija and R. B. Dubey
XGBoost: 2D-Object Recognition Using Shape Descriptors
and Extreme Gradient Boosting Classifier . . . . . . . . . . . . . . . . . . . . . . . 207
Monika, Munish Kumar, and Manish Kumar
Comparison of Principle Component Analysis and Stacked
Autoencoder on NSL-KDD Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Kuldeep Singh, Lakhwinder Kaur, and Raman Maini
Maintainability Configuration for Component-Based Systems
Using Fuzzy Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Kiran Narang, Puneet Goswami, and K. Ram Kumar
Development of Petri Net-Based Design Model for Energy Efficiency
in Wireless Sensor Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Sonal Dahiya, Ved Prakash, Sunita Kumawat, and Priti Singh
Lifting Wavelet and Discrete Cosine Transform-Based
Super-Resolution for Satellite Image Fusion . . . . . . . . . . . . . . . . . . . . . . 273
Anju Asokan and J. Anitha
Biologically Inspired Intelligent Machine and Its Correlation
to Free Will . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Munesh Singh Chauhan
Weather Status Prediction of Dhaka City Using Machine
Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Sadia Jamal, Tanvir Hossen Bappy, Roushanara Pervin,
and AKM Shahariar Azad Rabby
Contents ix

Image Processing: What, How and Future . . . . . . . . . . . . . . . . . . . . . . . 305

Mansi Lather and Parvinder Singh
A Study of Efficient Methods for Selecting Quasi-identifier
for Privacy-Preserving Data Mining . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
Rigzin Angmo, Veenu Mangat, and Naveen Aggarwal
Day-Ahead Wind Power Forecasting Using Machine Learning
Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
R. Akash, A. G. Rangaraj, R. Meenal, and M. Lydia
Query Relational Databases in Punjabi Language . . . . . . . . . . . . . . . . . 343
Harjit Singh and Ashish Oberoi
Machine Learning Algorithms for Big Data Analytics . . . . . . . . . . . . . . 359
Kumar Rahul, Rohitash Kumar Banyal, Puneet Goswami,
and Vijay Kumar
Fault Classification Using Support Vectors for Unmanned
Helicopters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Rupam Singh and Bharat Bhushan
EEG Signal Analysis and Emotion Classification
Using Bispectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Nelson M. Wasekar, Chandrkant J. Gaikwad, and Manoj M. Dongre
Slack Feedback Analyzer (SFbA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Ramchandra Bobhate and Jyoti Malhotra
A Review of Tools and Techniques for Preprocessing
of Textual Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
Abhinav Kathuria, Anu Gupta, and R. K. Singla
A U-Shaped Printed UWB Antenna with Three Band Rejection . . . . . . 423
Deepak Kumar, Preeti Rani, Tejbir Singh, and Vishant Gahlaut
Prediction Model for Breast Cancer Detection Using Machine
Learning Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Nishita Sinha, Puneet Sharma, and Deepak Arora
Identification of Shoplifting Theft Activity Through Contour
Displacement Using OpenCV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
Kartikeya Singh, Deepak Arora, and Puneet Sharma
Proof of Policy (PoP): A New Attribute-Based Blockchain
Consensus Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
R. Mythili and Revathi Venkataraman
x Contents

Real-Time Stabilization Control of Helicopter Prototype by IO-IPD

and L-PID Controllers Tuned Using Gray Wolf Optimization
Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
Hem Prabha, Ayush, Rajul Kumar, and Ankit Lal Meena
Factors of Staff Turnover in Textile Businesses in Colombia . . . . . . . . . 479
Erick Orozco-Acosta, Milton De la Hoz-Toscano, Luis Ortiz-Ospino,
Gustavo Gatica, Ximena Vargas, Jairo R. Coronado-Hernández,
and Jesus Silva
CTR Prediction of Internet Ads Using Artificial Organic Networks . . . . 489
Jesus Silva, Noel Varela, Danelys Cabrera,
and Omar Bonerge Pineda Lezama
Web Platform for the Identification and Analysis of Events
on Twitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
Amelec Viloria, Noel Varela, Jesus Vargas,
and Omar Bonerge Pineda Lezama
Method for the Recovery of Indexed Images in Databases
from Visual Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
Amelec Viloria, Noel Varela, Jesus Vargas,
and Omar Bonerge Pineda Lezama
Model for Predicting Academic Performance Through Artificial
Intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
Jesus Silva, Ligia Romero, Darwin Solano, Claudia Fernandez,
Omar Bonerge Pineda Lezama, and Karina Rojas
Feature-Based Sentiment Analysis and Classification
Using Bagging Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
Yash Ojha, Deepak Arora, Puneet Sharma, and Anil Kumar Tiwari
A Novel Image Encryption Method Based on LSB Technique
and AES Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
Paras Chaudhary
Implementing Ciphertext Policy Encryption in Cloud Platform
for Patients’ Health Information Based on the Attributes . . . . . . . . . . . 547
S. Boopalan, K. Ramkumar, N. Ananthi, Puneet Goswami,
and Suman Madan
Improper Passing and Lane-Change Related Crashes: Pattern
Recognition Using Association Rules Negative Binomial Mining . . . . . . 561
Subasish Das, Sudipa Chatterjee, and Sudeshna Mitra
Sleep Stage and Heat Stress Classification of Rodents Undergoing
High Environmental Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577
Prabhat Kumar Upadhyay and Chetna Nagpal
Contents xi

Development of a Mathematical Model for Solar Power Estimation

Using Regression Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
Arjun Viswanath, Karthik Krishna, T. Chandrika, Vavilala Purushotham,
and Priya Harikumar
Cloud Based Interoperability in Healthcare . . . . . . . . . . . . . . . . . . . . . . 599
Rakshit Joshi, Saksham Negi, and Shelly Sachdeva
Non-attendance of Lectures; Perceptions of Tertiary Students:
A Study of Selected Tertiary Institutions in Ghana . . . . . . . . . . . . . . . . 613
John Kani Amoako and Yogesh Kumar Sharma

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623

About the Editors

Dr. Vijendra Singh is working as Professor in the School of Computer Science at

The University of Petroleum and Energy Studies, Dehradun, Uttarakhand, India.
Prior to joining the UPES, he worked with the NCU, Delhi-NCR, India, Mody
University, Lakshmangarh, India, and Asian CERC Information Technology Ltd.
Dr. Singh received his Ph.D. degree in Engineering and M.Tech. degree in
Computer Science and Engineering from Birla Institute of Technology, Mesra,
India. He has 20 years of experience in research and teaching including IT industry.
Dr. Singh major research concentration has been in the areas of data mining, pattern
recognition, image processing, big data, machine learning, and soft computation.
He has published more than 65 scientific papers in this domain. He has served as
Editor-in-Chief, Special Issue, Procedia Computer Science, Vol 167, 2020,
Elsevier; Editor-in-Chief, Special Issue, Procedia Computer Science, Vol 132,
2018, Elsevier; Associate Editor, International Journal of Healthcare Information
Systems and Informatics, IGI Global, USA; Guest Editor, Intelligent Data Mining
and Machine Learning, International Journal of Healthcare Information Systems
and Informatics, IGI Global, USA; Editor-in-Chief, International Journal of Social
Computing and Cyber-Physical Systems, Inderscience, UK; Editorial Board
Member, International Journal of Multivariate Data Analysis, Inderscience, UK;
Editorial Board Member, International Journal of Information and Decision
Sciences, Inderscience, UK.

Dr. Vijayan K. Asari is a Professor in Electrical and Computer Engineering and

Ohio Research Scholars Endowed Chair in Wide Area Surveillance at the
University of Dayton, Dayton, Ohio. He is the Director of the University of Dayton
Vision Lab (Center of Excellence for Computer Vision and Wide Area Surveillance
Research). Dr. Asari had been a Professor in Electrical and Computer Engineering
at Old Dominion University, Norfolk, Virginia, till January 2010. He was the
Founding Director of the Computational Intelligence and Machine Vision
Laboratory (ODU Vision Lab) at ODU. Dr. Asari received the bachelor’s degree in
Electronics and Communication Engineering from the University of Kerala
(College of Engineering, Trivandrum), India, in 1978, the M.Tech. and Ph.D.

xiii
xiv About the Editors

degrees in Electrical Engineering from the Indian Institute of Technology, Madras,

in 1984 and 1994, respectively. Dr. Asari received several teachings, research,
advising, and technical leadership awards. Dr. Asari received the Outstanding
Teacher Award from the Department of Electrical and Computer Engineering in
April 2002 and the Excellence in Teaching Award from the Frank Batten College of
Engineering and Technology in April 2004. Dr. Asari has published more than 480
research papers including 80 peer-reviewed journal papers co-authoring with his
graduate students and colleagues in the areas of image processing, computer vision,
pattern recognition, machine learning, and high-performance digital system archi-
tecture design. Dr. Asari has been a Senior Member of the IEEE since 2001 and is a
Senior Member of the Society of Photo-Optical Instrumentation Engineers (SPIE).
He is a Member of the IEEE Computational Intelligence Society (CIS), IEEE CIS
Intelligent Systems Applications Technical Committee, IEEE Computer Society,
IEEE Circuits and Systems Society, Association for Computing Machinery (ACM),
and American Society for Engineering Education (ASEE).

Dr. Sanjay Kumar is working as Professor in the Computer Science and

Engineering Department, SRM University, India. He received his Ph.D. degree in
Computer Science and Engineering from Deenbandhu Chhotu Ram University of
Science and Technology (DCRUST), Murthal (Sonipat), in 2014. He obtained his
B.Tech. and M.Tech. degrees in Computer Science and Engineering in 1999 and
2005, respectively. He has more than 16 years of academic and administrative
experience. He has published more than 15 papers in the international and national
journals of repute. He has also presented more than 12 papers in the international
and national conferences. His current research area is wireless sensor networks,
machine learning, IoT, cloud computing, mobile computing and cyber, and network
security. He chaired the sessions in many international conferences like IEEE,
Springer, and Taylor & Francis. He is the Life Member of Computer Society of
India and Indian Society for Technical Education.

Prof. R. B. Patel is working as Professor in the Department of Computer Science

and Engineering, Chandigarh College of Engineering and Technology (CCET),
Chandigarh, India. Prior to joining the CCET, he worked as Professor at NIT,
Uttarakhand, India, and Dean, Faculty of Information Technology and Computer
Science, Deenbandhu Chhotu Ram University of Science and Technology, Murthal,
India. His research areas include mobile and distributed computing, machine and
deep learning, and wireless sensor networks. Prof. Patel has published more than
150 papers in international journals and conference proceedings. He has supervised
16 Ph.D. scholars and currently 02 are in progress.
Content Recommendation Based
on Topic Modeling

Sachin Papneja, Kapil Sharma, and Nitesh Khilwani

Abstract With the proliferation in Internet usage and communicating devices,

plenty amount of information is available at user disposal but on other side, it leads
to a challenge to provide the fruitful information to end users. To overcome this
problem, recommendation system plays a decisive role in providing pragmatic infor-
mation to end users at appropriate time. This paper proposes a topic modeling based
recommendation system to provide contents related to end users interest. Recom-
mendation systems are based on different filtering mechanisms which are classified
as content based, collaborative based, knowledge based, utility based and hybrid
filtering, etc. The objective of this research is thus to proffer a recommendation
system based on topic modeling. Benefit of latent Dirichlet allocation (LDA) is to
uncover latent semantic structure from the text documents. By analyzing the contents
using topic modeling, system can recommend the right articles to end users based
on user interest.

Keywords Recommendation system · LDA · Topic modeling · Content filtering ·

Collaborative filtering

1 Introduction

In last few years, with the telecom revolution, Internet has become a powerful tool
which has changed the way user communicate among themselves as well use it in
the professional business. As per year 2018 statistics, there are now more than 4

S. Papneja (B) · K. Sharma

Department of Computer Science & Engineering, Delhi Technological University,
New Delhi, India
e-mail: p.sachin@samsung.com
K. Sharma
e-mail: kapil@ieee.org
N. Khilwani
RoundGlass, New Delhi, India
e-mail: nitesh.nifft@gmail.com

© Springer Nature Singapore Pte Ltd. 2021 1

V. Singh et al. (eds.), Computational Methods and Data Engineering,
Advances in Intelligent Systems and Computing 1227,
https://doi.org/10.1007/978-981-15-6876-3_1
2 S. Papneja et al.

billion people around the world using the Internet whereas around 7.5 billion mobile
connections across the globe. As per the assessment, there are close to 1.5 billion
Internet sites on the cyberspace today. Out of the total available sites, less than 200
million are operating. As the number of communicating devices increase rapidly, it
results in the infinite amount of data in the embodiment of text, images, and videos.
Fundamental test is to give the exact data to the user dependent on user intrigue.
Amidst the appearance of Internet network accessibility, user’s propensities for
understanding news or most recent data have alternated from magazine or booklet
to advanced substance. Because of the immense amount of information accessible
on the cyberspace, it is extremely awkward for the end user to have the data acces-
sible according to his/her advantage. Recommender Systems aid conquers this issue
and gives important data or administrations to the user. Various sorts of suggestion
frameworks exist, for example, content based [17], collaborative [13], hybrid [7],
utility based, multi-criteria, context-aware, risk-aware based, each having with their
impediments. Analysts need to utilize distinctive suggestion frameworks dependent
on their exploration territories.
Content-based frameworks attempt to prescribe things like those a given user
has enjoyed before. For sure, the essential procedure performed by a content-based
recommender comprises in coordinating up the characteristics of a client profile in
which inclinations and interests are put away, with the properties of a substance
object (thing), so as to prescribe to the client new intriguing things. Content-based
recommenders exploit solely ratings. Content-based recommenders are capable of
recommending items not yet rated by any user provided by the active user to build
her own profile.
Numerous customary news recommender frameworks utilize collective sifting
to make suggestions dependent on the conduct of clients in the framework. In this
methodology, the presentation of new clients or new things can cause the cold start
issue, as there will be lacking information on these new sections for the communi-
tarian separating to draw any deductions for new clients or things. Content-based
news recommender frameworks developed to address the cold start issue. In any
case, many substance-based news recommender frameworks consider records as a
sack of-words disregarding the shrouded subjects of the news stories.
Individuals have consistently been standing up to with a developing measure of
information, which thusly requests more on their capacities to channel the substance
as indicated by their inclinations. Among the undeniably overpowering measures of
website pages, records, pictures, or recordings, it is never again natural to and what
we truly need. Besides, copy or a few data sources are discovered covering similar
themes. The clients are touchy to the recentness of data and their inclinations are
additionally changing after some time alongside the substance of the Web.
During the previous two decades, the ideas of recommender frameworks have
risen to cure the circumstance. The quintessence of recommender frameworks are
profoundly connected with the broad work in psychological science, guess hypoth-
esis, data recovery, determining speculations, and the board science. The content-
based methodology of suggestion has its foundations in data recovery [18], and
data separating [13] research. Content-based frameworks are planned for the most
Content Recommendation Based on Topic Modeling 3

part to suggest content-based things; the substance in these frameworks is generally

portrayed with keyword. Customized recommender frameworks intend to prescribe
applicable things to users dependent on their watched conduct, e.g., search person-
alization [3], Google News personalization [4], and Yahoo! conduct focusing on [5]
among others.
As of late, topic modeling approach, for example, latent Dirichlet allocation
(LDA) and probabilistic latent semantic analysis (pLSA) helps in examining content
substance includes by uncovering latent topics of each archive from the document
archive. Reason for LDA is to reveal the semantic structure covered up in the docu-
ments which incorporates the word appropriations over the latent subjects and the
inactive point disposal over archives [8]. The principle challenge is the way to suggest
explicit articles from an immense measure of recently accessible data to end customer,
where the chosen commodity should coordinate the buyer intrigue. In this research
work, concocted a recommendation framework dependent on LDA topic modeling.
In the recommended framework, LDA topic modeling is used to uncover topic from
the document related to user hobbies. Once system know the user interest, based on
the information system can recommend the articles related to interest by filtering
required information. One of the significant qualities of probabilistic topic modeling
is the capacity to uncover shrouded relations through the examination of co-event
designs on dyadic perceptions, for example, document-term pairs.

2 Related Researches

The main purpose of Recommender System is to assist users to make accurate deci-
sions without spending too much on searching this vast amount of information.
Traditional Recommender System is designed to recommend meaningful items to
their users. Those items depend on the purpose of the RS, for example, Google
recommends news to people while Facebook recommends people (friends) to people.
Recommender Systems are a sub-class of information retrieval systems and designed
to predict users’ future preferences by analyzing their past interaction with the system.
Usage of Recommender System became more common in recent years.
From the last two decades, Recommender Systems have become the topic of
interest for both academician and for the industry due to increase in overloaded infor-
mation and to provide relevant information to end users [1] by filtering out the infor-
mation. A knowledge-based filtering framework is a data framework intended for
unstructured or semi-organized information [5]. Recommender System may antici-
pate whether a end user would be keen on purchasing a specific item or not. Social
recommendation strategies gather assessment of commodity from numerous people,
and use nearest neighbor procedures to make proposals to a user concerning new
stock [4]. Recommendation system has been largely used in approximation theory
[14], cognitive science [16], forecasting theory, management science. In addition
to Recommender Systems works on the absolute values of ratings, [9] worked on
preference-based filtering, i.e., anticipating the general inclinations of end user.
4 S. Papneja et al.

Xia et al. [19] Suggested content-based recommender framework for E-Commerce

Platform and took a shot at streamlines the coupon picking process and customizes
the suggestion to improve the active clicking factor and, eventually, the conversion
rates. Deng et al. [10] proposed the amalgamation of item rating data that user has
given plus consolidated features of item to propose a novel recommendation model.
Bozanta and Kutlu [6] proposed to gathered client visit chronicles, scene related
data (separation, class, notoriety and cost) and relevant data (climate, season, date
and time of visits) identified with singular client visits from different sources as each
current scene suggestion framework calculation has its own disadvantages. Another
issue is that basic data about setting is not ordinarily utilized in scene suggestion
frameworks. Badriyah et al. [3] utilize proposed framework which suggest property-
related data based on the user action via looking through publicizing content recently
looked by the user. Topic modeling is based on the experience that document consist
of topics whereas topics are congregation of words. Goal of the Topic modeling is
to understand the documents by uncovering hidden latent variables which are used
to describe the document semantic. Latent Semantic Analysis is based on singular
value decomposition (SVD) whereas pLSA is based on probability distribution. LDA
is a Bayesian version of pLSA which uses Dirichlet priors for the document-topic
and word-topic distributions for better generalization.
Luostarinen and Kohonen [12] Studied and compared LDA with other standard
methods such as Naïve Bayes, K-nearest neighbor, regression and regular linear
regression and found out that LDA gives significant improvement in cold start
simulation.
Apaza et al. [2] use LDA by inferring topics from content given in a college course
syllabus for course recommendation to college students from sites such as Coursera,
Udacity, Edx, etc. Pyo et al. [15] proposed unified topic model for User Grouping and
TV program recommendation by employing two latent Dirichlet allocation (LDA)
models. One model is applied on TV users and the other on the viewed TV programs.

3 Background

3.1 Content-Based Recommender Systems

Content-Based (CB) Recommender Systems prescribe things to a user as indicated

by the substance of user’s past inclinations. As such, framework produces proposals
dependent on thing highlights that match with the user profile. The fundamental
procedure can be clarified in two primary advances:
1. Framework makes user profile utilizing user past conduct, all the more exactly
utilizing thing highlights that has been obtained or loved in the past by the user.
2. At that point, framework creates suggestion by breaking down the qualities of
these things and contrasting them and the user profile.
Content Recommendation Based on Topic Modeling 5

Content-based calculation can be comprehended from its name that this strategy for
the most part thinks about thing’s substance. Content-based strategy can be effectively
utilized in thing proposal; however, it necessitates that the applicable traits of the
things can be separated, or at the end of the day it depends on the thing’s substance.
For instance, on the off chance that framework prescribes archives to its users, at that
point the content-based calculation examines reports’ words (content). Be that as it
may, a few things’ highlights cannot be removed effectively, for example, motion
pictures and music, or they can be covered because of security issues consequently
materialness of these techniques is constrained relying upon the idea of the things.
Probabilistic topic models are a suite of methods whose objective is to detect the
concealed topical structure in enormous chronicles of documents.

3.2 Recommender Systems Major Challenges

There are numerous difficulties that recommender framework researchers face today
and those difficulties can influence the algorithm outcome. Some of the challenges
are as follows:
• Data sparsity: Nowadays a great many things are accessible particularly in
online business sites and every day this number is expanding. Along these lines,
finding comparative user (that purchased comparative things) is getting more
enthusiastically. A large portion of the Recommender System calculations are
utilizing user/things closeness to create recommenders. Along these lines, due to
information sparsity calculations may not perform precisely.
• Scalability: Especially, enormous sites have a large number of user and a great
many information. In this way, when planning a Recommender System it ought
to likewise think about the computational expense.
• Cold Start: When new user or information enter the system, system cannot draw
any data hence it cannot produce proposals either. One of the most guileless
answers for the cold start issue is prescribing well known or stylish things to new
users. For instance, in YouTube, when a user has no past video history story it
will prescribe the most famous recordings to this user. In any case, when the user
watches a video then system will have some clue regarding the client’s inclination
and afterward it will prescribe comparative recordings to the past video that the
client has viewed.
• Diversity and accuracy: It is typically viable to prescribe famous things to users.
In any case, users can likewise discover those things independent from anyone else
without a recommender framework. Recommender framework ought to likewise
locate the less famous things however are probably going to be favored by the users
to suggest. One answer for this issue is utilizing mixture suggestion techniques.
• Vulnerability to attacks: Recommender Systems can be focus of a few assaults
attempting to mishandle the Recommender System calculations utilized in the
6 S. Papneja et al.

web-based business sites. Those assaults attempt to trick Recommender System

to wrongly propose foreordained things for benefit.
• The value of time: Customer needs/inclinations will in general change in time.
Be that as it may, most Recommender Systems calculations don’t think about time
as a parameter.
• Evaluation of recommendations: There are a few Recommender System struc-
tured with various purposes and measurements proposed to assess the Recom-
mender System. Notwithstanding, how to pick the one that precisely assesses the
comparing framework is as yet not clear.

3.3 Probabilistic Topic Modeling

Today there are a large amount of articles, site pages, books and web journals acces-
sible on the web. Besides, every day the measure of content reports are expanding
with commitments from informal communities and mechanical improvements. In
this way, finding what we are actually searching for is not a simple assignment as it
used to be and it tends to be very tedious. For instance, for researchers, there are a
million of articles accessible on the web, to locate the related ones is a challenging
task for researchers. It is not practical to peruse every content and compose or classify
them. Along these lines, it is important to utilize programming devices to sort out
them.
For instance, most journals chronicle their issues, putting away every distributed
article, and along these lines, they should store a lot of data. Without utilizing compu-
tational devices arranging such a major unstructured text assortment is unimaginable
by just utilizing human work. In this way, researchers evolve distinctive probabilistic
models for subject revelation from an enormous unstructured text corpus and they
called them probabilistic topic models.
Probabilistic subject models are calculations intended to find the concealed topic
of the article. At the end of the day, they are measurable techniques attempting to
find the shrouded topic of each article by breaking down the recurrence of the words.
The primary thought behind theme models is a presumption that articles are blends
of points (ordinary dispersion) and subjects are typical circulation over words. Topic
models are generative models which fundamentally imply that producing a document
is considered as a probabilistic procedure. This procedure can be clarified in three
fundamental points as pursues:
• Determine an article to be produced.
• Pick topic for every word of the article.
• Draft a word dependent on the topic that has been picked.
Despite the fact that theme models are initially intended to arrange or locate
the shrouded subject of unstructured archives, they have been embraced in a wide
range of spaces with various sorts of information. For instance, they are used in data
retrieval, multimedia retrieval.
Content Recommendation Based on Topic Modeling 7

Probabilistic Topic Modeling comes under the non-supervised learning [11] in

the sense that it does not require antecedent interpretation or document labeling.
In probabilistic modeling, information is exuded from a generative procedure that
incorporates latent variables. This generative procedure characterizes a joint prob-
ability distribution over both the noticed and concealed random variables. It does
not make any earlier supposition how the words are showed up in the document yet
rather what is important to the model is the occurrence of the word is referenced in
the document.

3.4 Latent Dirichlet Allocation

Latent Dirichlet Allocation (LDA) is a three-level hierarchical Bayesian model, in

which every collected item is demonstrated as a limited blend over a basic arrange-
ment of topics and is utilized to reveal topics in a lot of documents. Every topic is, thus,
demonstrated as a limitless blend over a hidden arrangement of topic probabilities.
Document is only having some data about the topic while every topic is portrayed by
dissemination over words. The LDA model is spoken to as a probabilistic graphical
model as shown in Fig. 1.
As it tends to be seen from the diagram that there are three unique degrees of
factors and parameters:
• First level is corpus level parameters and they are examined in the first place for
example before start producing the corpus.
• Second level is record level factors and they are tested once for producing each
archive.
• Third level factors are word-level factors and they are created for each expression
of all records in the corpus.

N
α θ z W
M

Fig. 1 LDA graphical model

8 S. Papneja et al.

In Fig. 1, document is described by M though each document is succession of N

words where word is signified by w and topic variable in document is characterized
by z. The parameters α and β are corpus-level parameters and are inspected once
during the time spent creating a corpus. The factors θ is document level variable,
examined once per document. Lastly, the factors z and ware word-level factors and
are examined once for each word in each document.

4 Proposed System

To provide contents related to user interest, each article related to interest is consid-
ered as document. LDA is used to find out the semantic structure concealed in the
document. LDA provided us a topic distribution for each interest area, so this learning
will help to recommend the related article to end user based on the user interest. LDA
consider each document as collection of topics in a certain distribution and each topic
as a collection of keywords. Once number of topics is decided as input to LDA algo-
rithm, it firstly rearranges the topic proportion within the document and keyword
distribution with in a topic to have a good configuration of topic-keyword.
Accuracy of LDA algorithm depends on some key factors:
1. Quality of input text.
2. Number and variety of topics.
3. Tuning parameters.
In our experiment, we have taken three different topics (cooking, cricket and
bodybuilding) as a user interest for an input to LDA algorithm. Data is gathered
from different websites by writing a crawler in python. Before inputting the data to
the LDA algorithm. All collected data has been cleaned by removing stop words,
removing emails, remove new line characters and remove distracting single quotes.
Once data is preprocessed, now all the sentences are converted into words. To have
more accuracy build the bigram model and performed the lemmatization on the words
followed by removing words whose count is either less than 15% or more than 50%
of the words. Now corpus will be created. Now the preprocessed data is separated
into training set and test set. Once model is prepared with the training set, model
accuracy is checked using the test data.
In Fig. 2, all the three topics are well segregated and have a keywords weight age
for all the three topics.

5 Conclusion and Future Scope

In this paper, content recommendation based on topic modeling is studied and imple-
mented. Implementation is performed on python by considering document related
to three topics and accuracy achieved is 89%. In the future, work will be extended
Content Recommendation Based on Topic Modeling 9

Fig. 2 Topic 1 most relevant terms

by considering more number of different topic documents and system will provide
personalized content to the end users.

References

1. Adomavicius G, Tuzhilin A (2005) Toward the next generation of recommender systems: a

survey of the state-of-the-art and possible extension. Ieee Trans Knowl Data Eng 17(6):734–749
2. Apaza RG, Cervantes EV, Quispe LC, Luna JO (2014) Online courses recommendation based
on LDA. In: Symposium on information management and big data—SIMBig 2014. Peru, p 7
3. Badriyah T, Azvy S, Yuwono W, Syarif I (2018) Recommendation system for property
search using content based filtering method. In: International Conference on Information and
Communications Technology. Yogyakarta
4. Basu C, Hirsh H, Cohen W (1998) Recommendation as classification: using social and content-
based information in recommendation. Am Assoc Artif Intell, p 7. (USA)
5. Belkin NJ, Croft WB (1992). Information filtering and information retrieval: two sides of the
same coin? Commun ACM 35(12):29–38
6. Bozanta A, Kutlu B (2018) HybRecSys: content-based contextual hybrid venue recommender
system. J Inf Sci 45(2)
7. Burke R (2007) Hybrid web recommender systems. Springer-Verlag, Berlin
8. Chang TM, Hsiao W-F (2013) LDA-based personalized document recommendation. Pacific
Asia Conf Inf Sys 13
9. Cohen WW, Schapire RE, Singer Y (1999) Learning to order things. J Artif Intell Res 10:243–
270
10. Deng F, Ren P, Qin Z, Huang G, Qin Z (2018) August). Leveraging image visual features in
content-based recommender system, Hindawi Scientific Programming, p 8
11. Duda RO, Hart PE, Stork DG (2001) Pattern classification. Wiley
12. Luostarinen T, Kohonen O (2013) Using topic models in content-based news recommender
systems. In: 19th Nordic conference of computational linguistics. Oslo, Norway, p 13
10 S. Papneja et al.

13. Pazzani MJ, Billsus D (2007) Content-based recommendation systems. The Adaptive Web.
Berlin, pp 325–341
14. Powell M (1981) Approximation theory and methods. In: Press CU (ed) Press Syndicate of the
University of Cambridge, New York, USA
15. Pyo S, Kim M, Kim E (2014) LDA-based unified topic modeling for similar TV user grouping
and TV program recommendation. IEEE Trans Cybern 16
16. Rich E (1979) User modeling via stereotypes. Elsevier 3(4):329–354
17. Sarkhel JK, Das P (2010) Towards a new generation of reading habits in Internet Era. In: 24th
national seminar of IASLIC. Gorakhpur University, U.P, pp 94–102
18. Su X, Khoshgoftaar TM (2009) A survey of collaborative filtering techniques. Adv Artif Intell
2009(421425), 19
19. Xia Y, Fabbrizio GD, Vaibhav S, Datta A (2017) A content-based recommender system for
e-commerce offers and coupons. SIGIR eCom. Tokyo, p 7
Hybrid ANFIS-GA and ANFIS-PSO
Based Models for Prediction of Type 2
Diabetes Mellitus

Ratna Patil, Sharvari Tamane, and Nirmal Rawandale

Abstract Type- Diabetes Mellitus (T2DM), a major threat to developing as well

as developed countries, can be easily controlled to a large extent through lifestyle
modifications. Diabetes increases the risk of developing various health as well as
financial problems to cure these health complications. The health complications are
stroke, myocardial infarction, and coronary artery disease. Nerve, muscle, kidney
and retinal damage have distressing impact on the life of a diabetic patient. It is
the need of the hour to halt the epidemic of T2DM in the early stage. Data science
approaches have the potential to predict on medical data. Machine learning is an
evolving scientific field in data science where machines learn mechanically and
improve from experience without any explicit program. Our goal was to develop a
system which can improve performance of a classifier for prediction of T2DM. The
purpose of this work is to implement a hybrid model for prediction by integrating the
advantages of artificial neural net (ANN) and fuzzy logic. Genetic algorithm (GA)
and particle swarm optimization (PSO) have been applied to optimize parameters of
developed predicting model. The proposed scheme used a fuzzification matrix. This
matrix is used to relate the input patterns with a degree of membership to different
classes. The specific class is predicted based on the value of degree of membership
of a pattern. We have analyzed the proposed method and previous research in the
literature. High accuracy was achieved using the ANFIS-PSO approach.

Keywords Machine learning · Fuzzy system · Diabetes mellitus · Particle swarm

intelligence approach · Adaptive neuro-fuzzy inference system (ANFIS)

R. Patil (B)
Noida Institute of Engineering and Technology, Greater Noida, Uttar Pradesh, India
e-mail: ratna.nitin.patil@gmail.com
S. Tamane
Jawaharlal Nehru Engineering College, Aurangabad, India
e-mail: sharvaree73@yahoo.com
N. Rawandale
Shri Bhausaheb Hire Government Medical College, Dhule, India
e-mail: drrawandale@gmail.com

© Springer Nature Singapore Pte Ltd. 2021 11

V. Singh et al. (eds.), Computational Methods and Data Engineering,
Advances in Intelligent Systems and Computing 1227,
https://doi.org/10.1007/978-981-15-6876-3_2
12 R. Patil et al.

1 Introduction

Diabetes Mellitus is classified into three types. These are namely Type-I (T1DM),
Type-II (T2DM), and Gestational DM (GDM). T2DM appears to be the most
common form of diabetes in India where more than one crore cases are reported
per year. It is developed if insulin is not produced adequately by the pancreas. The
main contributing factors of T2DM include lifestyle, physical inactivity, obesity,
eating habits, and genetics. In T2DM human body does not use insulin properly. We
have considered T2DM for our study. Several classification algorithms are designed
for classifying the patients as diabetic or healthy. ANFIS has its place in the class of
hybrid structure, termed as neuro-fuzzy systems. ANFIS receives the properties of
neural net as well as fuzzy systems [1]. Neural networks can learn effortlessly from
the input provided but it is hard to understand the knowledge assimilated through
neural net [2]. In contrast, fuzzy-based models are understood very straightforwardly.
Fuzzy inference system (FIS) exploits linguistic terms instead of numeric values and
generates rules in the form of if-then structure. Linguistic variables have values in the
form of words in natural language having degrees of membership. Partial member-
ship is allowed in fuzzy sets, which shows that an element exists in more than one
set partially.
The usage of ANFIS makes the creation of the rule base more adaptive to the state
for modeling and controlling of complex and non-linear problems. In this approach,
the rule base is created by exploiting the neural network systems through the back-
propagation process. To boost its performance, the properties of fuzzy logic are
inherited in this model.
In the proposed method, the fusion of ANFIS with metaheuristic approach
has been done. Metaheuristic algorithm follows repetitive process. Metaheuristic
methods control a subordinate heuristic by exploiting and exploring the search space.
These algorithms are stimulated by seeing the phenomena happening in the nature.
This paper is systematized as follows: Related work done by other researchers
is discussed in Sect. 2. Section 3 includes discussion and construction of ANFIS
process. Discussion on GA is represented in Sect. 4 and PSO is depicted in Sect. 5.
Section 6 presents the building of proposed algorithm. Experimental results are
discussed and results obtained are compared in Sect. 7. Lastly, in Sect. 8 concluding
remarks are made.

2 Related Work

ANFIS has been used commonly as an effective tool for prediction due to its learning
abilities and this approach facilitates rapid adaptation to deviations in systems which
directed to robust groundwork for research. In this background work done by other
researchers is presented here.
Hybrid ANFIS-GA and ANFIS-PSO Based Models … 13

Author Soumadip Ghosh have has analyzed the performance of three different
techniques NFS, RBFNN, and ANFIS widely used in Data Mining [3]. Performance
was analyzed based on root mean square error (RMSE), Kappa statistic, F-measure,
accuracy percentage on ten standard datasets from UCI. The results suggest that
ANFIS has RMSE value of 0.4205.
Author Alby in his paper has developed ANFIS with GA and General Regression
Neural Network (GRNN) for prediction of Type-II DM [4]. Using ANFIS with GA
accuracy was 93.49% and accuracy was 85.49% with GRNN classifier.
Authors Ratna, Sharvari Tamne have done the comparison and analysis of logistic
regression (LR), decision tree, K nearest neighbors (KNN), gradient boost, Gaussian
Naïve Bayes, MLP, support vector machine (SVM), and random forest algorithms
[5]. In this study, they have stated the strength and limitations of existing work.
Author Sinan Adnan Diwan Alalwan has carried out a detailed literature survey
on different methods for predicting T2DM [6]. In his work, he has suggested random
forest method and self-organizing map for improving the accuracy of prediction.
Several authors have used PCA technique for dimensionality reduction of dataset.
Authors Ratna et al. have used PCA for dimensionality reduction technique followed
by KMeans in their study and have shown that performance was improved [7, 8].
Author Murat et al. used PCA followed by ANFIS for diagnosing diabetes [9].
Author Quan Zou has implemented three classifiers using random forest, decision
tree, and neural network methods. He has analyzed and compared these classifiers
on PIMA and Luzhou dataset [10]. The study shows that random forests are better
than the other two. For dimensionality reduction PCA and minimum redundancy
maximum relevance (mRMR) were employed. But the result shows that accuracy
was 0.8084 which was better when all the features were used with random forest.
Authors Patil and Tamane have developed the genetic algorithm for feature selec-
tion with K nearest neighbor (KNN) and Naïve Bayes approach [11]. Though both
the models have improved the accuracy of the prediction with reduced feature set,
GA + KNN have got the better results than GA + Naïve Bayes. In GA + KNN
approach, validation accuracy has been improved from 74% to 83%.

3 ANFIS

ANFIS is a fuzzy inference system introduced by Jang, 1993. It is implemented in

the framework of adaptive systems. ANFIS architecture is depicted in Fig. 1. ANFIS
network has two membership functions. Inputs are converted to fuzzy values using
input membership function. Generally used input membership functions are Trian-
gular, Trapezoidal, Gaussian. Fuzzy output of FIS is mapped to crisp value by output
membership functions. Tuning of parameters related with the membership function
is completed during the learning phase. Gradient vector is used for computation of
these parameters and their tuning. For a specified set of parameters, gradient vector
actually computes a measure of how fine the FIS has modeled the provided data.
After getting the gradient vector one of various optimization method can be used
14 R. Patil et al.

Fig. 1 5-layered architecture of ANFIS

for adjusting the parameters for minimizing error measure. This degree of error is
generally calculated by the sum of the squared difference between actual and wanted
outputs.
For approximation of membership function parameters, ANFIS employs either
back-propagation or combination least squares estimation with back-propagation.
Fuzzy rules are created using Sugeno-type fuzzy system on a specified dataset. A
typical form of Sugeno fuzzy rule is:
IF I 1 is Z 1
AND I 2 is Z 2
.....
AND I m is Z m
THEN y = f (I 1 , I 2 ,…, I m )
Where,
I 1 , I 2 ,…, I m are input variables;
Z 1 , Z 2 ,…, Z m are fuzzy sets.
There are five layers with different functions in ANFIS architecture. These
layers are called as fuzzification, product, normalization, de-fuzzy, and output layer
sequentially. Equations (1) to (6) depict function of each layer.
Layer 1: It is a fuzzy layer where the crisp signal is given as input to the ith node.
This node is linked with a linguistic label Ai or else Bi−2 . The function computes
the membership value of the input. The input layer calculates the output from all the
nodes by applying Eqs. (1) and (2).

O1, i = μ Ai (X ), where i = 1, 2 (1)

Hybrid ANFIS-GA and ANFIS-PSO Based Models … 15

O1, i = μ Bi−2 (Y ), where i = 3, 4 (2)

In Eqs. (1) and (2) the inputs to ith node are given by X, Y and Ai , Bi are representing
linguistic symbols. μAi is the membership function of Ai .
Layer 2: All the nodes in this product layer are fixed nodes characterized as . A
rule neuron computes firing strength W i by the product of all the incoming signals
by Eq. (3). Each node output implies the firing strength of a rule.

O2,i = Wi = min {μ Ai (X ), μ Bi (Y )}, where i = 1, 2 (3)

Layer 3: Every node in normalization layer calculates normalized firing strength of

a given rule. It is proportion of the firing strength of specified rule to the summation
of firing strengths of all rules. It indicates the involvement of a given rule to the
ultimate result. Consequently, the output from ith neuron in layer 3 is calculated by
Eq. (4).

Wi
O3,i = Wi = , where i = 1, 2 (4)
(W1 + W2 )

Layer 4: Each neuron in the defuzzification layer computes the weighted

consequential value of a certain rule by Eq. (5).

O4,i = Wi f i = Wi ( pi x + qi y + ri ), where i = 1, 2 (5)


Layer 5: The output layer has a single node. This is a fixed node having label .
It computes the overall ANFIS output by adding the outputs from all the neurons in
the defuzzification layer as in Eq. (6).

O5,1 = Wi f i (6)
i

4 Genetic Algorithm (GA)

They are generally used to produce solutions for optimization and exploration tasks.
GA simulates “survival of-the-fittest” between individuals of succeeding genera-
tion for problem-solving. Genetic algorithms use methods inspired by evolutionary
biology such as selection, inheritance, alteration, and recombination. Pseudocode of
GA is given below:
1. Select initial population.
2. Compute the fitness of every candidate in the populace.
3. Repetition of the next steps (a–e) until termination condition is satisfied.
16 R. Patil et al.

a. High ranking entities are selected for reproduction.

b. Use recombination operator to yield next generation.
c. The resultant offspring is mutated.
d. Evaluate the offspring.
e. Substitution of low ranked chunk of populace with the reproduced descend
ants.

5 Particle Swarm Optimization (PSO)

Kennedy and Eberhart developed PSO in 1995. It is a stochastic optimization method.

The concept of PSO is analogous to flight of birds in hunt of food. It is an evolutionary
optimization method built on the movement and intellect of swarms [12]. PSO is
a population-based searching process where swarm of particles are the searching
agents and position of particle gives solution. Each particle is considered to be a
point (candidate solution) in a N-dimensional space which fine-tunes its “flying”
based on its personal flying experience and the flying experience of other particles.
This concept is represented in Fig. 2.
PSO has found its way in modeling of biological and sociological behavior like
group of birds looking for food cooperatively. The PSO has been also extensively
used in population-based hunt approach. In a search space, the position of particle
is changed repetitively until it reaches to the best solution or until the computational
boundaries are reached.
Pseudocode of PSO is mentioned below:

Fig. 2 PSO concept

Another Random Scribd Document
with Unrelated Content
palme d’épaisseur, et qu’il soit fait d’os doublé d’acier au dedans et
au dehors.
Si la lance de Roger avait pu supporter ce rude choc, et si, au
premier coup, elle ne s’était pas rompue en mille morceaux qui
volèrent jusqu’au ciel comme s’ils eussent eu des ailes, elle aurait
percé le haubert, ce dernier eût-il été plus dur que le diamant, et le
combat aurait été fini. Mais elle se rompit. Les deux destriers allèrent
toucher la terre avec leur croupe.
Cependant les cavaliers relèvent promptement leurs destriers de
la bride et de l’éperon ; jetant leurs lances, ils tirent leur épée, et
reviennent l’un sur l’autre pleins de fureur et de rage. Faisant
caracoler de côté et d’autre, avec beaucoup d’adresse, leurs chevaux
dociles et légers, ils cherchent de la pointe de l’épée le défaut de
leurs cuirasses.
La poitrine de Rodomont n’était plus protégée par la rude écaille
du serpent ; il n’avait plus à la main l’épée tranchante de Nemrod, et
son front n’était plus armé de son casque ordinaire. Il avait laissé les
armes qu’il portait d’habitude, suspendues au monument d’Isabelle,
après avoir été vaincu sur le pont par la dame de Dordogne, comme
il me semble vous l’avoir dit plus haut.
Il avait une nouvelle armure fort bonne, mais qui était loin d’être
aussi parfaite que la première. Mais pas plus l’ancienne que la
nouvelle n’aurait arrêté Balisarde, à laquelle ne résistait ni
enchantement, ni finesse d’acier, ni dureté de trempe. Roger
s’escrime si bien de çà et de là, qu’il a déjà percé les armes du païen
en plus d’un endroit.
Quand le païen voit son sang rougir ses armes de tous côtés, et
qu’il ne peut éviter que la plus grande partie des coups qu’on lui
porte arrivent jusqu’à sa chair, il est saisi d’une rage plus grande,
d’une fureur plus intense que la mer un jour de tempête au cœur de
l’hiver. Il jette son écu, et prenant son épée à deux mains, il frappe
de toutes ses forces sur le casque de Roger.
La machine qui est supportée sur le Pô par deux bateaux, et dont
le marteau relevé au moyen d’hommes et de roues, retombe sur les
poutres aiguisées en pointes, ne frappe pas des coups plus
formidables que celui que le fier païen asséna de toutes ses forces
sur la tête de Roger. Ce dernier fut protégé par son casque
enchanté ; sans cela, lui et son cheval auraient été fendus d’un seul
coup.
Roger s’incline à deux reprises ; il ouvre les bras et les jambes
comme s’il allait tomber. Avant qu’il ait eu le temps de se remettre,
le Sarrasin lui porte un second coup plus terrible, suivi d’un
troisième. Mais son glaive trop faible ne peut supporter une si rude
besogne ; il vole en éclats, et laisse la main du cruel païen désarmée.
Rodomont ne s’arrête point pour cela. Il s’approche de Roger qui
est encore privé de sentiment, tellement les coups qu’il a reçus sur
la tête lui ont troublé la cervelle. Mais le Sarrasin ne tarde pas à le
réveiller de ce lourd sommeil ; de son bras puissant, il lui enlace le
cou et le serre avec une telle force, qu’il l’enlève des arçons, et le
jette à terre.
Roger n’a pas plus tôt touché la terre, qu’il se redresse plein de
colère et de vergogne. Il jette les yeux sur Bradamante. Il la voit si
troublée de sa chute, que son beau visage pâlit et que la vie est
prête à l’abandonner. Roger, désireux d’effacer promptement cette
honte que Rodomont lui a fait subir, saisit son épée et fond sur le
païen.
Celui-ci le heurte de son destrier, mais Roger l’esquive
adroitement en se rejetant en arrière. Au moment où le destrier
passe devant lui, il le saisit à la bride de la main gauche, et le force à
tourner sur lui-même, tandis que, de la main droite, il cherche à
frapper le cavalier soit au flanc, soit au ventre, soit à la poitrine. Il
finit par lui porter deux coups de pointe, l’une au flanc, l’autre à la
cuisse.
Rodomont, qui tenait encore à la main le pommeau de son épée
brisée, en assène un tel coup sur le casque de Roger, qu’il aurait dû
l’étourdir de nouveau. Mais Roger qui devait vaincre, ayant le bon
droit pour lui, le saisit par le bras, et joignant sa main droite à la
première, tire son adversaire tant et si bien, qu’il finit par l’arracher
de selle.
Soit force, soit adresse, le païen tombe de façon qu’il n’a plus
d’avantage sur Roger ; je veux dire qu’il retombe à pied. Mais Roger
qui a encore son épée, est mieux partagé. Il s’efforce de tenir le
païen à distance, afin d’éviter une lutte corps à corps avec un
adversaire d’une taille si gigantesque.
Il voit le sang couler de son flanc, de sa cuisse et de ses autres
blessures. Il espère que, peu à peu, les forces lui manqueront, et
qu’il pourra achever de le vaincre. Le païen avait encore à la main le
pommeau de son épée ; réunissant toutes ses forces, il en porte un
coup qui étourdit Roger plus qu’il ne le fut jamais.
Roger, frappé à la visière de son casque et à l’épaule, vacille et
chancelle sous le coup, et a toutes les peines du monde à se tenir
debout. Le païen veut s’élancer sur lui, mais le pied lui manque,
affaibli qu’il est par sa blessure à la cuisse. Dans sa précipitation à
s’élancer sur Roger, il tombe sur un genou.
Roger ne perd pas de temps ; il lui porte de grands coups à la
poitrine et à la figure ; il le martelle, et le tient en respect en le
maintenant à terre avec la main. Mais le païen fait si bien, qu’il
réussit à se relever ; il saisit Roger, et l’enlace dans ses bras. L’un et
l’autre, joignant l’adresse à la force, cherche à ébranler, à étouffer
son adversaire.
Rodomont, blessé à la cuisse, et le flanc ouvert, avait perdu une
grande partie de ses forces. Roger, depuis longtemps rompu à tous
les exercices du corps, possédait une grande adresse. Il comprend
son avantage et ne s’en dessaisit pas. La où il voit le sang sortir avec
le plus d’abondance des blessures du païen, il pèse de tout le poids
de ses bras, de sa poitrine, de ses deux pieds.
Rodomont, plein de rage et de dépit, a saisi Roger par le cou et
par les épaules. Il le tire, il le secoue, il le soulève de terre et le tient
suspendu sur sa poitrine. Il le serre étroitement, l’ébranle de çà de
là, et cherche à le faire tomber. Roger, ramassé sur lui-même, fait
appel à toute son adresse, à toute sa vigueur, pour garder
l’avantage.
Le franc et brave Roger finit par saisir Rodomont. Il pèse avec sa
poitrine sur le flanc droit de son adversaire, et le serre de toute ses
forces ; en même temps, il lui passe la jambe droite sous le genou
gauche, tandis que son autre jambe enlace la jambe de Rodomont.
Il le soulève ainsi de terre, et le renverse la tête la première.
Rodomont va frapper le sol de la tête et des épaules. La secousse
est si violente, que le sang jaillit de ses blessures comme de deux
fontaines, et rougit au loin la terre. Roger qui sent que la Fortune est
pour lui, redouble d’efforts. Afin d’empêcher le Sarrasin de se
relever, il lui porte d’une main le poignard à la visière, de l’autre il le
tient à la gorge ; avec ses genoux, il lui presse le ventre.
Parfois, dans les mines d’or de la Pannonie ou de l’Ibérie, un
éboulement subit vient ensevelir ceux que leur avarice y a fait
descendre ; les malheureux sont tellement étouffés, que leur souffle
peut à peine s’exhaler. Il en est de même du Sarrasin, oppressé sous
le poids de son vainqueur et renversé par terre.
Roger a tiré son poignard ; il en porte la pointe à la visière de
Rodomont et lui crie de se rendre, en lui promettant de lui laisser la
vie. Mais celui-ci, qui redoute moins de mourir que de montrer un
seul instant de faiblesse, s’agite, se secoue et, sans répondre,
cherche à mettre Roger sous lui.
De même qu’un mâtin, renversé par un dogue féroce qui lui a
enfoncé ses crocs dans la gorge, s’agite et se débat en vain, les yeux
ardents et la gueule baveuse, et ne peut se débarrasser de son
redoutable adversaire qui le surpasse en force mais non en rage,
ainsi le païen finit par perdre tout espoir de se délivrer de l’étreinte
de Roger victorieux.
Cependant, il se tord et se débat de telle sorte qu’il réussit à
dégager son bras droit et à tirer son poignard. Il cherche à frapper
Roger sous les reins ; mais le jeune homme s’aperçoit du danger
qu’il court s’il tarde plus longtemps à donner la mort à cet
indomptable Sarrasin.
 Levant le bras le plus qu’il peut, il plonge deux ou trois fois tout
entier le fer de son poignard dans le front horrible de Rodomont, et
se dégage ainsi de tout péril. Vers les affreuses rives d’Achéron,
délivrée du corps plus froid que glace, s’enfuit, en blasphémant,
l’âme dédaigneuse qui fut si altière et si orgueilleuse au monde.

FIN DU TOME QUATRIÈME ET DERNIER

 NOTES DU TOME QUATRIÈME

CHANT XXXVII.

[1] Page 2, ligne 27. — Arpalice. — Fille du roi de Thrace, qui

défendit le royaume paternel contre Pyrrhus, fils d’Achille.
[2] Page 2, ligne 27. — Tomyris. — Reine des Massagètes, dont
Hérodote raconte la victoire sur Cyrus, roi de Perse.
[3] Page 2, ligne 27. — Celle qui secourut Turnus. — Camille, fille
du roi des Volsques, qui aida Turnus contre Énée. (Énéide, liv. 10 et
11.)
[4] Page 2, ligne 28. — Celle qui vint en aide à Hector. —
Pentésilée, reine des Amazones.
[5] Page 2, lignes 28, 29 et 30. — Celle qui, suivie des gens de
Sidon et de Tyr, et longeant le rivage d’Afrique, s’établit en Libye. —
Didon, fondatrice de Carthage.
[6] Page 2, ligne 30. — Zénobie. — Reine de Palmyre, faite
prisonnière par l’empereur Aurélien, après avoir opposé une longue
résistance.
[7] Page 2, ligne 31. — Celle qui sauva par ses victoires les
Assyriens, les Perses et les Indiens. — Sémiramis.
[8] Page 5, lignes 14 et 15. — S’abreuver avec les Muses à la
fontaine d’Aganippe. — Cette fontaine jaillissait sur l’Hélicon, et était
consacrée à Apollon et aux Muses. Ses eaux avaient la vertu
d’inspirer les poètes.
[9] Page 24, lignes 16 et 17. — Le grand fleuve qui sort du
Vésale. — Le Pô qui descend du mont Viso, dans les Alpes
liguriennes.

CHANT XXXVIII.

[10] Page 43, ligne 28. — Le désert où l’armée de Cambyse fut

détruite. — Cambyse, roi de Perse, ayant envoyé une armée contre
les Ammons, peuple de la Libye, vers les confins de la Cyrénaïque,
cette armée fut engloutie par les sables soulevés par le vent du
désert. (Hérodote.)

CHANT XL.

[11] Page 78, lignes 5 et 6. — Le jour où vous lui montrâtes la

flotte ennemie écrasée à l’embouchure du Pô. — Allusion à la défaite
des Vénitiens sur le Pô, par le cardinal d’Este.
[12] Page 85, lignes 19 et 20. — S’ouvre un âpre sentier dans les
champs d’Ocnus. — Le poète entend par là, la campagne autour de
Mantoue, parce que cette ville avait eu, croyait-on, pour fondateur
Ocnus, fils de la nymphe Manto.
[13] Page 88, lignes 6 et 7. — Ludovic le More, remis aux mains
d’un autre Ludovic. — Ludovic le More, trahi par les Suisses qu’il
avait pris à son service, fut livré à Louis XII, roi de France.

CHANT XLII.

[14] Page 146, ligne 17. — Par ce Camille, dont le Reno et

Felsina écoutent les chants. — Camille Paleotto, de Bologne, attaché
au cardinal de Bibbiena, et qui, de concert avec Guido Postumo,
dont il est parlé quelques lignes plus loin, chanta les louanges de
Lucrezia Bentivogli, fille naturelle du duc de Ferrare.
 CHANT XLIII.

[15] Page 165, lignes 23 et 24. — Personne ne voudrait plus

entendre parler de la patrie de Nausica. — L’île de Corfou,
renommée dans l’antiquité pour la beauté des jardins d’Alcinoüs,
père de Nausica.

CHANT LXV.

[16] Page 229, lignes 3, 4 et 5. — Servius, Marius et Ventidius

l’ont montré dans l’antiquité, et, de notre temps, le roi Louis. — Le
roi de France Louis XII.

FIN DES NOTES DU TOME QUATRIÈME

 TABLE DES MATIÈRES DU TOME
QUATRIÈME

Pages.

ROLAND FURIEUX

Chant XXXVII. — Passant en revue les écrivains divers qui

ont employé leur plume à chanter les louanges du beau
sexe, le poète en prend occasion pour louer Vittoria
Colonna et les nobles vers consacrés par elle à la
mémoire du marquis de Pescaire, son époux. Puis il
introduit sur la scène Ullania, messagère de la reine de
l’île Perdue, qui raconte à Roger, à Bradamante et à
Marphise l’indigne coutume établie par Marganor dans
son propre château à l’encontre des femmes. Les deux
guerrières et Roger infligent à Marganor le châtiment
qu’il a mérité. 1
Chant XXXVIII. — Roger, fidèle à l’honneur qui l’appelle 33
auprès d’Agramant, s’en va à Arles. Bradamante et
Marphise se présentent à la cour de Charles. Marphise
reçoit le baptême. — Astolphe, à la tête d’une armée
de Nubiens, saccage l’Afrique et menace Biserte.
Agramant, instruit de ces événements, obtient de
 Charles de décider de la guerre entre eux par le combat
singulier de deux champions élus dans chaque camp.
Chant XXXIX. — Mélisse, au moyen d’un enchantement, fait
qu’Agramant viole le pacte juré. Les deux armées en
viennent aux mains, et les Maures ont le dessous. —
Astolphe accomplit des prouesses en Afrique, et y crée
une flotte. Ses compagnons et lui s’emparent de
Roland, et Astolphe lui rend la raison. — Agramant,
s’étant embarqué avec ses troupes, rencontre la flotte
chrétienne qui l’attaque. 56
Chant XL. — La flotte d’Agramant ayant été battue et
brûlée, les chrétiens assiègent Biserte, qui est prise
d’assaut, mise au pillage et livrée aux flammes.
Agramant se réfugie à Lampéduse avec Sobrin. Ayant
trouvé Gradasse dans cette île, ils arrêtent tous les trois
le projet de défier Roland et deux autres chevaliers
chrétiens au combat. Roland accueille volontiers cette
offre, et choisit pour compagnons Brandimart et Olivier.
— Entre temps, Roger, retournant à Arles, délivre sept
rois africains que Dudon conduisait prisonniers, et en
vient aux mains avec ce dernier. 77
Chant XLI. — Roger et Dudon cessent leur combat, après
être convenus que les sept rois païens prisonniers
seront rendus à la liberté. Roger s’embarque avec eux
pour l’Afrique. Pendant la traversée, ils sont engloutis
par une tempête, excepté Roger, qui est porté sain et
sauf près d’un ermite, lequel lui prédit diverses choses.
— Le navire, abandonné par son équipage, est poussé
par le vent jusqu’à Biserte. Il y avait à bord l’épée,
l’armure et le cheval de Roger. Roland prend l’épée
pour lui, donne l’armure à Olivier et le cheval à
Brandimart, et ils vont tous les trois à Lampéduse pour
combattre les trois païens. Le combat s’engage ; Sobrin
et Olivier sont blessés, et Brandimart est tué. 98
Chant XLII. — Le combat de Lampéduse se termine par la
mort de Gradasse et d’Agramant, occis par la main de
Roland, qui accorde la vie à Sobrin. — Bradamante se
désole du retard de Roger. — Renaud, en allant sur les
traces d’Angélique, trouve un remède qui le guérit de
son amoureuse passion. S’étant remis en chemin pour
rejoindre Roland, il fait la rencontre d’un chevalier qui le
reçoit dans un magnifique palais orné de statues
représentant diverses dames de la maison d’Este. Son
hôte lui propose un moyen de s’assurer de la fidélité de
sa femme. 124
Chant XLIII. — Renaud entend raconter deux nouvelles,
l’une contre les femmes, l’autre contre les hommes qui
se laissent vaincre par l’ignoble passion de l’avarice.
Après un long chemin sur terre et sur mer, Renaud
arrive à Lampéduse, au moment où venait de se
terminer le combat entre les paladins et les païens. Ils
descendent tous en Sicile et, sur la plage d’Agrigente,
ils rendent les derniers honneurs aux dépouilles
mortelles de Brandimart. De là, ils vont à l’ermitage où
est Roger, devenu déjà chrétien. L’ermite rend la santé
à Olivier et à Sobrin, qui se fait aussi baptiser. 151
Chant XLIV. — Les cinq guerriers se lient d’une fraternelle 202
amitié. Renaud, tenant Roger en grande estime, et sur
les conseils de l’ermite, lui promet la main de sa sœur
Bradamante. De là, ils s’en vont à Marseille, où arrive
en même temps Astolphe, qui a licencié son armée de
Nubiens, et rendu sa flotte à son premier état de
feuilles. Les paladins et Sobrin sont magnifiquement
accueillis par Charles dans Paris, mais la joie générale
est troublée par le refus du duc Aymon et de sa femme
Béatrice de consentir à l’union de Roger et de
Bradamante, celle-ci ayant été déjà fiancée par eux à
Léon, fils de l’empereur des Grecs. Roger prend ses
armes et, plein de haine contre Léon, il se transporte
 au camp des Bulgares qui sont en guerre avec les
Grecs. Il défait ces derniers, puis va loger dans une
hôtellerie qu’il ignore être située sur les terres de
l’empire grec. Il y est dénoncé comme l’auteur du
désastre éprouvé par les Grecs.
Chant XLV. — Roger, saisi pendant son sommeil, devient le
prisonnier de Théodora, sœur de l’empereur
Constantin. — Entre temps, Charles, à la requête de
Bradamante, a fait publier que quiconque voudra l’avoir
pour femme devra se battre avec elle et la vaincre. —
Léon, qui a conçu de l’amitié et de l’estime pour Roger,
sans le connaître, le tire de prison et l’engage à
combattre en son nom contre Bradamante. Roger,
portant les insignes de Léon, se bat contre la guerrière.
Survient la nuit ; Charles fait cesser le combat et donne
Bradamante à celui qu’il croit être Léon. Roger,
désespéré, veut se tuer ; mais Marphise va trouver
Charles et empêche ce mariage. 228
Chant XLVI. — Le poète, se sentant arriver au port, nomme
les nombreux amis qui l’attendent pour fêter son retour.
— Mélisse va à la recherche de Roger, et lui sauve la vie
avec le concours de Léon qui, ayant appris le motif du
désespoir de Roger, lui cède Bradamante. Tous vont à
Paris, où Roger, élu déjà roi des Hongrois, est reconnu
pour le chevalier qui a combattu contre Bradamante.
On célèbre les noces avec une splendeur royale ; le lit
nuptial est préparé sous la tente impériale que Mélisse,
grâce à son art magique, a fait venir de Constantinople.
Pendant le dernier jour des fêtes, survient Rodomont
qui défie Roger ; le combat a lieu, et Rodomont reçoit la
mort de la main de Roger. 258
Notes 294
FIN DE LA TABLE DES MATIÈRES DU TOME QUATRIÈME ET DERNIER.
 ACHEVÉ D’IMPRIMER
LE TRENTE AVRIL MIL HUIT CENT QUATRE-VINGT-UN
PAR A. QUANTIN
POUR
ALPHONSE LEMERRE, ÉDITEUR
A PARIS
*** END OF THE PROJECT GUTENBERG EBOOK ROLAND FURIEUX,
TOME 4 ***

Updated editions will replace the previous one—the old editions will
be renamed.

Creating the works from print editions not protected by U.S.

copyright law means that no one owns a United States copyright in
these works, so the Foundation (and you!) can copy and distribute it
in the United States without permission and without paying
copyright royalties. Special rules, set forth in the General Terms of
Use part of this license, apply to copying and distributing Project
Gutenberg™ electronic works to protect the PROJECT GUTENBERG™
concept and trademark. Project Gutenberg is a registered trademark,
and may not be used if you charge for an eBook, except by following
the terms of the trademark license, including paying royalties for use
of the Project Gutenberg trademark. If you do not charge anything
for copies of this eBook, complying with the trademark license is
very easy. You may use this eBook for nearly any purpose such as
creation of derivative works, reports, performances and research.
Project Gutenberg eBooks may be modified and printed and given
away—you may do practically ANYTHING in the United States with
eBooks not protected by U.S. copyright law. Redistribution is subject
to the trademark license, especially commercial redistribution.

START: FULL LICENSE

THE FULL PROJECT GUTENBERG LICENSE
 PLEASE READ THIS BEFORE YOU DISTRIBUTE OR USE THIS WORK

To protect the Project Gutenberg™ mission of promoting the free

distribution of electronic works, by using or distributing this work (or
any other work associated in any way with the phrase “Project
Gutenberg”), you agree to comply with all the terms of the Full
Project Gutenberg™ License available with this file or online at
www.gutenberg.org/license.

Section 1. General Terms of Use and

Redistributing Project Gutenberg™
electronic works
1.A. By reading or using any part of this Project Gutenberg™
electronic work, you indicate that you have read, understand, agree
to and accept all the terms of this license and intellectual property
(trademark/copyright) agreement. If you do not agree to abide by all
the terms of this agreement, you must cease using and return or
destroy all copies of Project Gutenberg™ electronic works in your
possession. If you paid a fee for obtaining a copy of or access to a
Project Gutenberg™ electronic work and you do not agree to be
bound by the terms of this agreement, you may obtain a refund
from the person or entity to whom you paid the fee as set forth in
paragraph 1.E.8.

1.B. “Project Gutenberg” is a registered trademark. It may only be

used on or associated in any way with an electronic work by people
who agree to be bound by the terms of this agreement. There are a
few things that you can do with most Project Gutenberg™ electronic
works even without complying with the full terms of this agreement.
See paragraph 1.C below. There are a lot of things you can do with
Project Gutenberg™ electronic works if you follow the terms of this
agreement and help preserve free future access to Project
Gutenberg™ electronic works. See paragraph 1.E below.
1.C. The Project Gutenberg Literary Archive Foundation (“the
Foundation” or PGLAF), owns a compilation copyright in the
collection of Project Gutenberg™ electronic works. Nearly all the
individual works in the collection are in the public domain in the
United States. If an individual work is unprotected by copyright law
in the United States and you are located in the United States, we do
not claim a right to prevent you from copying, distributing,
performing, displaying or creating derivative works based on the
work as long as all references to Project Gutenberg are removed. Of
course, we hope that you will support the Project Gutenberg™
mission of promoting free access to electronic works by freely
sharing Project Gutenberg™ works in compliance with the terms of
this agreement for keeping the Project Gutenberg™ name associated
with the work. You can easily comply with the terms of this
agreement by keeping this work in the same format with its attached
full Project Gutenberg™ License when you share it without charge
with others.

1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside the
United States, check the laws of your country in addition to the
terms of this agreement before downloading, copying, displaying,
performing, distributing or creating derivative works based on this
work or any other Project Gutenberg™ work. The Foundation makes
no representations concerning the copyright status of any work in
any country other than the United States.

1.E. Unless you have removed all references to Project Gutenberg:

1.E.1. The following sentence, with active links to, or other

immediate access to, the full Project Gutenberg™ License must
appear prominently whenever any copy of a Project Gutenberg™
work (any work on which the phrase “Project Gutenberg” appears,
or with which the phrase “Project Gutenberg” is associated) is
accessed, displayed, performed, viewed, copied or distributed:
 This eBook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this eBook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.

1.E.2. If an individual Project Gutenberg™ electronic work is derived

from texts not protected by U.S. copyright law (does not contain a
notice indicating that it is posted with permission of the copyright
holder), the work can be copied and distributed to anyone in the
United States without paying any fees or charges. If you are
redistributing or providing access to a work with the phrase “Project
Gutenberg” associated with or appearing on the work, you must
comply either with the requirements of paragraphs 1.E.1 through
1.E.7 or obtain permission for the use of the work and the Project
Gutenberg™ trademark as set forth in paragraphs 1.E.8 or 1.E.9.

1.E.3. If an individual Project Gutenberg™ electronic work is posted

with the permission of the copyright holder, your use and distribution
must comply with both paragraphs 1.E.1 through 1.E.7 and any
additional terms imposed by the copyright holder. Additional terms
will be linked to the Project Gutenberg™ License for all works posted
with the permission of the copyright holder found at the beginning
of this work.

1.E.4. Do not unlink or detach or remove the full Project

Gutenberg™ License terms from this work, or any files containing a
part of this work or any other work associated with Project
Gutenberg™.

1.E.5. Do not copy, display, perform, distribute or redistribute this

electronic work, or any part of this electronic work, without
prominently displaying the sentence set forth in paragraph 1.E.1
with active links or immediate access to the full terms of the Project
Gutenberg™ License.

1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if you
provide access to or distribute copies of a Project Gutenberg™ work
in a format other than “Plain Vanilla ASCII” or other format used in
the official version posted on the official Project Gutenberg™ website
(www.gutenberg.org), you must, at no additional cost, fee or
expense to the user, provide a copy, a means of exporting a copy, or
a means of obtaining a copy upon request, of the work in its original
“Plain Vanilla ASCII” or other form. Any alternate format must
include the full Project Gutenberg™ License as specified in
paragraph 1.E.1.

1.E.7. Do not charge a fee for access to, viewing, displaying,

performing, copying or distributing any Project Gutenberg™ works
unless you comply with paragraph 1.E.8 or 1.E.9.

1.E.8. You may charge a reasonable fee for copies of or providing

access to or distributing Project Gutenberg™ electronic works
provided that:

• You pay a royalty fee of 20% of the gross profits you derive
from the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
 about donations to the Project Gutenberg Literary Archive
Foundation.”

• You provide a full refund of any money paid by a user who

notifies you in writing (or by e-mail) within 30 days of receipt
that s/he does not agree to the terms of the full Project
Gutenberg™ License. You must require such a user to return or
destroy all copies of the works possessed in a physical medium
and discontinue all use of and all access to other copies of
Project Gutenberg™ works.

• You provide, in accordance with paragraph 1.F.3, a full refund of

any money paid for a work or a replacement copy, if a defect in
the electronic work is discovered and reported to you within 90
days of receipt of the work.

• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.

1.E.9. If you wish to charge a fee or distribute a Project Gutenberg™

electronic work or group of works on different terms than are set
forth in this agreement, you must obtain permission in writing from
the Project Gutenberg Literary Archive Foundation, the manager of
the Project Gutenberg™ trademark. Contact the Foundation as set
forth in Section 3 below.

1.F.

1.F.1. Project Gutenberg volunteers and employees expend

considerable effort to identify, do copyright research on, transcribe
and proofread works not protected by U.S. copyright law in creating
the Project Gutenberg™ collection. Despite these efforts, Project
Gutenberg™ electronic works, and the medium on which they may
be stored, may contain “Defects,” such as, but not limited to,
incomplete, inaccurate or corrupt data, transcription errors, a
copyright or other intellectual property infringement, a defective or
damaged disk or other medium, a computer virus, or computer
codes that damage or cannot be read by your equipment.

1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES - Except for

the “Right of Replacement or Refund” described in paragraph 1.F.3,
the Project Gutenberg Literary Archive Foundation, the owner of the
Project Gutenberg™ trademark, and any other party distributing a
Project Gutenberg™ electronic work under this agreement, disclaim
all liability to you for damages, costs and expenses, including legal
fees. YOU AGREE THAT YOU HAVE NO REMEDIES FOR
NEGLIGENCE, STRICT LIABILITY, BREACH OF WARRANTY OR
BREACH OF CONTRACT EXCEPT THOSE PROVIDED IN PARAGRAPH
1.F.3. YOU AGREE THAT THE FOUNDATION, THE TRADEMARK
OWNER, AND ANY DISTRIBUTOR UNDER THIS AGREEMENT WILL
NOT BE LIABLE TO YOU FOR ACTUAL, DIRECT, INDIRECT,
CONSEQUENTIAL, PUNITIVE OR INCIDENTAL DAMAGES EVEN IF
YOU GIVE NOTICE OF THE POSSIBILITY OF SUCH DAMAGE.

1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If you

discover a defect in this electronic work within 90 days of receiving
it, you can receive a refund of the money (if any) you paid for it by
sending a written explanation to the person you received the work
from. If you received the work on a physical medium, you must
return the medium with your written explanation. The person or
entity that provided you with the defective work may elect to provide
a replacement copy in lieu of a refund. If you received the work
electronically, the person or entity providing it to you may choose to
give you a second opportunity to receive the work electronically in
lieu of a refund. If the second copy is also defective, you may
demand a refund in writing without further opportunities to fix the
problem.

1.F.4. Except for the limited right of replacement or refund set forth
in paragraph 1.F.3, this work is provided to you ‘AS-IS’, WITH NO
OTHER WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.

1.F.5. Some states do not allow disclaimers of certain implied

warranties or the exclusion or limitation of certain types of damages.
If any disclaimer or limitation set forth in this agreement violates the
law of the state applicable to this agreement, the agreement shall be
interpreted to make the maximum disclaimer or limitation permitted
by the applicable state law. The invalidity or unenforceability of any
provision of this agreement shall not void the remaining provisions.

1.F.6. INDEMNITY - You agree to indemnify and hold the Foundation,

the trademark owner, any agent or employee of the Foundation,
anyone providing copies of Project Gutenberg™ electronic works in
accordance with this agreement, and any volunteers associated with
the production, promotion and distribution of Project Gutenberg™
electronic works, harmless from all liability, costs and expenses,
including legal fees, that arise directly or indirectly from any of the
following which you do or cause to occur: (a) distribution of this or
any Project Gutenberg™ work, (b) alteration, modification, or
additions or deletions to any Project Gutenberg™ work, and (c) any
Defect you cause.

Section 2. Information about the Mission

of Project Gutenberg™
Project Gutenberg™ is synonymous with the free distribution of
electronic works in formats readable by the widest variety of
computers including obsolete, old, middle-aged and new computers.
It exists because of the efforts of hundreds of volunteers and
donations from people in all walks of life.

Volunteers and financial support to provide volunteers with the

assistance they need are critical to reaching Project Gutenberg™’s
goals and ensuring that the Project Gutenberg™ collection will
remain freely available for generations to come. In 2001, the Project
Gutenberg Literary Archive Foundation was created to provide a
secure and permanent future for Project Gutenberg™ and future
generations. To learn more about the Project Gutenberg Literary
Archive Foundation and how your efforts and donations can help,
see Sections 3 and 4 and the Foundation information page at
www.gutenberg.org.

Section 3. Information about the Project

Gutenberg Literary Archive Foundation
The Project Gutenberg Literary Archive Foundation is a non-profit
501(c)(3) educational corporation organized under the laws of the
state of Mississippi and granted tax exempt status by the Internal
Revenue Service. The Foundation’s EIN or federal tax identification
number is 64-6221541. Contributions to the Project Gutenberg
Literary Archive Foundation are tax deductible to the full extent
permitted by U.S. federal laws and your state’s laws.

The Foundation’s business office is located at 809 North 1500 West,

Salt Lake City, UT 84116, (801) 596-1887. Email contact links and up
to date contact information can be found at the Foundation’s website
and official page at www.gutenberg.org/contact

Section 4. Information about Donations to

the Project Gutenberg Literary Archive
Foundation
Project Gutenberg™ depends upon and cannot survive without
widespread public support and donations to carry out its mission of
increasing the number of public domain and licensed works that can
be freely distributed in machine-readable form accessible by the
widest array of equipment including outdated equipment. Many
small donations ($1 to $5,000) are particularly important to
maintaining tax exempt status with the IRS.

The Foundation is committed to complying with the laws regulating

charities and charitable donations in all 50 states of the United
States. Compliance requirements are not uniform and it takes a
considerable effort, much paperwork and many fees to meet and
keep up with these requirements. We do not solicit donations in
locations where we have not received written confirmation of
compliance. To SEND DONATIONS or determine the status of
compliance for any particular state visit www.gutenberg.org/donate.

While we cannot and do not solicit contributions from states where

we have not met the solicitation requirements, we know of no
prohibition against accepting unsolicited donations from donors in
such states who approach us with offers to donate.

International donations are gratefully accepted, but we cannot make

any statements concerning tax treatment of donations received from
outside the United States. U.S. laws alone swamp our small staff.

Please check the Project Gutenberg web pages for current donation
methods and addresses. Donations are accepted in a number of
other ways including checks, online payments and credit card
donations. To donate, please visit: www.gutenberg.org/donate.

Section 5. General Information About

Project Gutenberg™ electronic works
Professor Michael S. Hart was the originator of the Project
Gutenberg™ concept of a library of electronic works that could be
freely shared with anyone. For forty years, he produced and
distributed Project Gutenberg™ eBooks with only a loose network of
volunteer support.
Project Gutenberg™ eBooks are often created from several printed
editions, all of which are confirmed as not protected by copyright in
the U.S. unless a copyright notice is included. Thus, we do not
necessarily keep eBooks in compliance with any particular paper
edition.

Most people start at our website which has the main PG search
facility: www.gutenberg.org.

This website includes information about Project Gutenberg™,

including how to make donations to the Project Gutenberg Literary
Archive Foundation, how to help produce our new eBooks, and how
to subscribe to our email newsletter to hear about new eBooks.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com