Nirmalya Chowdhury
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Nirmalya Chowdhury has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Debaditya Barman
Department of Computer and System Sciences, Visva-Bharati University, Santiniketan 731235, India
15 shared publications Ritam Sarkar
Department of Computer Science and Engineering, Jadavpur University, Kolkata 700032, India
4 shared publications- Anil Tudu
University of North Bengal
2 shared publications Anindya Bhattacharya
2 shared publicationsC.A. Murthy
2 shared publicationsRajat K. De
2 shared publications
Latest Publications
- 29 June 2023
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 83 (4) , 11461-11491
A cooperative co-evolutionary genetic algorithm for query recommendation
- 29 June 2023
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 83 (4) , 11461-11491
A cooperative co-evolutionary genetic algorithm for query recommendation
- 6 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Social Systems
- Vol. 7 (3) , 802-817
Overlapping Community Detection Using Multiobjective Genetic Algorithm
Community detection or identification problem has emerged as an important task in the field of network analysis. The problem becomes more challenging when the communities overlap each other. Since the community detection problem can be modeled as an optimization problem, the genetic algorithm (GA) can be used to develop a solution. Typically, overlapping community detection problem is a multiobjective optimization problem and there may exist some conflicts among these objectives. Therefore, a simple GA-based solution is not efficient for detecting overlapping communities in a social network. For this purpose, we have proposed an overlapping community detection method based on nondominated sorting GA (NSGA)-II having two objective functions. The first objective function maximizes internal link density within the community using fuzzy membership values while the second objective function minimizes the external link density of the communities. The initial population has been created based on the neighbors. A mutation operator has been proposed in this article to restrict the randomness in the chromosome. An updation operator has also been proposed to modify a chromosome after crossover and mutation operation. The proposed method can handle different network structures properly. In order to show the effectiveness, the proposed method has been compared with some popular methods on several real-world environments in terms of two performance measurement techniques, namely, generalized normalized mutual information (gNMI) and extended modularity (Qov). The empirical results reveal that our proposed method has succeeded to obtain better gNMI and comparable extended modularity values.
- 6 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Social Systems
- Vol. 7 (3) , 802-817
Overlapping Community Detection Using Multiobjective Genetic Algorithm
Community detection or identification problem has emerged as an important task in the field of network analysis. The problem becomes more challenging when the communities overlap each other. Since the community detection problem can be modeled as an optimization problem, the genetic algorithm (GA) can be used to develop a solution. Typically, overlapping community detection problem is a multiobjective optimization problem and there may exist some conflicts among these objectives. Therefore, a simple GA-based solution is not efficient for detecting overlapping communities in a social network. For this purpose, we have proposed an overlapping community detection method based on nondominated sorting GA (NSGA)-II having two objective functions. The first objective function maximizes internal link density within the community using fuzzy membership values while the second objective function minimizes the external link density of the communities. The initial population has been created based on the neighbors. A mutation operator has been proposed in this article to restrict the randomness in the chromosome. An updation operator has also been proposed to modify a chromosome after crossover and mutation operation. The proposed method can handle different network structures properly. In order to show the effectiveness, the proposed method has been compared with some popular methods on several real-world environments in terms of two performance measurement techniques, namely, generalized normalized mutual information (gNMI) and extended modularity (Qov). The empirical results reveal that our proposed method has succeeded to obtain better gNMI and comparable extended modularity values.
- 17 February 2020
- Published by Taru Publications in Journal of Interdisciplinary Mathematics
- Vol. 23 (2) , 523-535
Personalized query recommendation system : A genetic algorithm approach
Search engine has become an integral part of our daily life. It helps users to find a specific desirable information from the large amount of data stored in the web. Query recommendation feature of the search engine recommends several alternate queries to the user in response to the user’s original query. These recommendations essentially stay same for different users from a specific geographical region but acceptability of these alternate queries may vary from person to person. A genetic algorithm based personalized recommendation system has been proposed in this work. The proposed method has been evaluated using a commercial search engine’s search logs.
- 17 February 2020
- Published by Taru Publications in Journal of Interdisciplinary Mathematics
- Vol. 23 (2) , 523-535
Personalized query recommendation system : A genetic algorithm approach
Search engine has become an integral part of our daily life. It helps users to find a specific desirable information from the large amount of data stored in the web. Query recommendation feature of the search engine recommends several alternate queries to the user in response to the user’s original query. These recommendations essentially stay same for different users from a specific geographical region but acceptability of these alternate queries may vary from person to person. A genetic algorithm based personalized recommendation system has been proposed in this work. The proposed method has been evaluated using a commercial search engine’s search logs.
- 7 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Social Systems
- Vol. 6 (6) , 1190-1205
$k$ -Context Technique: A Method for Identifying Dense Subgraphs in a Heterogeneous Information Network
Most real-life problems can be modeled using heterogeneous networks, as they consist of several interconnected entities. However, researchers often study most of these problems as projected homogeneous networks. In such networks, different entities can be merged, some entities can be removed, or interactions among some entities can be overlooked, which results in a significant information loss. Moreover, these networks often overlook different structural possibilities in heterogeneous networks. Different structural possibilities may include a hierarchical network topology or a centralized network topology. Mining on different structural topologies of a heterogeneous information network may lead to a better understanding of the behavior and characteristics of the individual entities and the network as a whole. This article proposes a novel approach named k-context technique for extracting dense subgraphs from different structural topologies of the heterogeneous information network. The approach has been evaluated using both synthetic and real-life data sets, and it compared the results with an existing algorithm for finding dense subgraphs from a homogeneous information network.
- 7 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Social Systems
- Vol. 6 (6) , 1190-1205
$k$ -Context Technique: A Method for Identifying Dense Subgraphs in a Heterogeneous Information Network
Most real-life problems can be modeled using heterogeneous networks, as they consist of several interconnected entities. However, researchers often study most of these problems as projected homogeneous networks. In such networks, different entities can be merged, some entities can be removed, or interactions among some entities can be overlooked, which results in a significant information loss. Moreover, these networks often overlook different structural possibilities in heterogeneous networks. Different structural possibilities may include a hierarchical network topology or a centralized network topology. Mining on different structural topologies of a heterogeneous information network may lead to a better understanding of the behavior and characteristics of the individual entities and the network as a whole. This article proposes a novel approach named k-context technique for extracting dense subgraphs from different structural topologies of the heterogeneous information network. The approach has been evaluated using both synthetic and real-life data sets, and it compared the results with an existing algorithm for finding dense subgraphs from a homogeneous information network.
- 18 August 2019
- Published by Springer Nature in Advances in Intelligent Systems and Computing
- p. 727-740
A Cooperative Co-evolutionary Genetic Algorithm for Multi-Robot Path Planning Having Multiple Targets
- 18 August 2019
- Published by Springer Nature in Advances in Intelligent Systems and Computing
- p. 727-740
A Cooperative Co-evolutionary Genetic Algorithm for Multi-Robot Path Planning Having Multiple Targets
- 8 June 2018
- Published by Springer Nature
- p. 259-268
To Detect the Influencers in a Dynamic Co-authorship Network Using Heat-Diffusion Model
- 8 June 2018
- Published by Springer Nature
- p. 259-268
To Detect the Influencers in a Dynamic Co-authorship Network Using Heat-Diffusion Model
- 21 April 2018
- Published by Springer Nature
- p. 255-266
A Nonnegative Matrix Factorization Based Approach to Extract Aspects from Product Reviews
- 21 April 2018
- Published by Springer Nature
- p. 255-266
A Nonnegative Matrix Factorization Based Approach to Extract Aspects from Product Reviews
- 20 April 2018
- Published by Springer Nature in International Journal of Information Technology
- Vol. 12 (4) , 1147-1157
A novel semi supervised approach for text classification
- 20 April 2018
- Published by Springer Nature in International Journal of Information Technology
- Vol. 12 (4) , 1147-1157
A novel semi supervised approach for text classification
- 4 February 2016
- Published by Springer Nature
- p. 143-151
Classification of Bank Direct Marketing Data Using Subsets of Training Data
- 4 February 2016
- Published by Springer Nature
- p. 143-151
Classification of Bank Direct Marketing Data Using Subsets of Training Data
- 4 September 2015
- Published by Springer Nature
- p. 711-724
Opinion Classification Based on Product Reviews from an Indian E-Commerce Website
- 4 September 2015
- Published by Springer Nature
- p. 711-724
Opinion Classification Based on Product Reviews from an Indian E-Commerce Website