Soumyabrata Saha
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Soumyabrata Saha has a total of 55 co-authors. Here are the top co-authors by number of shared publications:
- Suparna DasGupta
Department of Information Technology, JIS College of Engineering, Kalyani, West Bengal, India
21 shared publications Rituparna Chaki
5 shared publications- Rituparna Chaki
University of Calcutta
4 shared publications Suman Kumar Das
Zenlabs, Zensar Technologies, Pune, 411014, Maharashtra, India
4 shared publications- Amrut Ranjan Jena
Department of Computer Science and Engineering, Dr. Sudhir Chandra Sur Institute of Technology and Sports Complex, Kolkata, West Bengal, India
3 shared publications - Anirban Bhar
Department of Computer Science and Engineering, University of Calcutta, Kolkata, India
3 shared publications
Latest Publications
- 23 May 2026
- Published by SAGE Publications in Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology
TriFusionRank: An AI Based Multi-Factor Sentence Scoring Framework for Lightweight Extractive Summarization
Extractive summarization is the key to alleviating the problem of information overload by choosing essential sentences from a document and creating short summaries. In this paper, a new extractive summarization method that combines three different scoring mechanisms, including term frequency-inverse document frequency (TF-IDF), sentence positional weighting, and semantic similarity with the document title, is proposed. In contrast to traditional techniques based on statistical or graph-based models, our method utilizes a hybrid approach to best capture sentence salience and contextual utility. Experimental evaluations on benchmark data show that the proposed method has competitive ROUGE scores against strong baselines while being interpretable and of low computational cost. This method provides a light but efficient solution for real-time summarization operations and provides a basis for future developments that include semantic comprehension and deep learning integration.
- 23 May 2026
- Published by SAGE Publications in Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology
TriFusionRank: An AI Based Multi-Factor Sentence Scoring Framework for Lightweight Extractive Summarization
Extractive summarization is the key to alleviating the problem of information overload by choosing essential sentences from a document and creating short summaries. In this paper, a new extractive summarization method that combines three different scoring mechanisms, including term frequency-inverse document frequency (TF-IDF), sentence positional weighting, and semantic similarity with the document title, is proposed. In contrast to traditional techniques based on statistical or graph-based models, our method utilizes a hybrid approach to best capture sentence salience and contextual utility. Experimental evaluations on benchmark data show that the proposed method has competitive ROUGE scores against strong baselines while being interpretable and of low computational cost. This method provides a light but efficient solution for real-time summarization operations and provides a basis for future developments that include semantic comprehension and deep learning integration.
- Open Access
- 19 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 28775-28784
A Domain-Specific Framework for Priority-Based Modeling and Optimization of Non-Functional Requirements Beyond Binary Constraints
Non-Functional Requirements (NFRs) such as performance, security, reliability, and usability significantly influence the overall quality of software systems. However, traditional NFR engineering approaches commonly treat these requirements as binary constraints, i.e., either fully satisfied or not, which limits systematic trade-off reasoning even in conventional single-system software architectures. This limitation becomes critical in complex, dynamic domains such as IoT and cyber-physical systems, where multiple interdependent NFRs must coexist without degrading system performance. To address this challenge, this paper proposes a domain-specific framework for modeling and optimizing priority-based NFRs beyond binary constraints. The framework integrates standard well-recognized quality models like ISO/IEC 25010 and FURPS+, to identify domain-specific NFRs, which are then modeled as soft goals with graded degrees of satisfaction. Priority weights, utility-based reasoning, and multi-objective optimization are integrated to manage conflicts and trade-offs among competing NFRs while preserving system stability. Unlike traditional approaches that enforce a rigidly defined subset of NFRs, the proposed framework enables the combination of multiple NFRs proportionally according to their priority and contextual relevance. The framework is applicable to diverse software-intensive systems including IoT, cloud platforms, healthcare systems, and cyber-physical environments. This work contributes a systematic, scalable, and adaptive approach to proactive NFR management suitable for modern software systems. The concrete research problem addressed is the failure of binary NFR handling in dynamic adaptation settings, leading to over-constrained designs or degraded quality. The framework guarantees improved global utility (15-25% in simulations) and adaptation success (>90%) while maintaining performance thresholds, with low runtime overhead (~O(n2) for n NFRs).
- 19 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 28775-28784
Open AccessA Domain-Specific Framework for Priority-Based Modeling and Optimization of Non-Functional Requirements Beyond Binary Constraints
Non-Functional Requirements (NFRs) such as performance, security, reliability, and usability significantly influence the overall quality of software systems. However, traditional NFR engineering approaches commonly treat these requirements as binary constraints, i.e., either fully satisfied or not, which limits systematic trade-off reasoning even in conventional single-system software architectures. This limitation becomes critical in complex, dynamic domains such as IoT and cyber-physical systems, where multiple interdependent NFRs must coexist without degrading system performance. To address this challenge, this paper proposes a domain-specific framework for modeling and optimizing priority-based NFRs beyond binary constraints. The framework integrates standard well-recognized quality models like ISO/IEC 25010 and FURPS+, to identify domain-specific NFRs, which are then modeled as soft goals with graded degrees of satisfaction. Priority weights, utility-based reasoning, and multi-objective optimization are integrated to manage conflicts and trade-offs among competing NFRs while preserving system stability. Unlike traditional approaches that enforce a rigidly defined subset of NFRs, the proposed framework enables the combination of multiple NFRs proportionally according to their priority and contextual relevance. The framework is applicable to diverse software-intensive systems including IoT, cloud platforms, healthcare systems, and cyber-physical environments. This work contributes a systematic, scalable, and adaptive approach to proactive NFR management suitable for modern software systems. The concrete research problem addressed is the failure of binary NFR handling in dynamic adaptation settings, leading to over-constrained designs or degraded quality. The framework guarantees improved global utility (15-25% in simulations) and adaptation success (>90%) while maintaining performance thresholds, with low runtime overhead (~O(n2) for n NFRs).
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 132938-132950
Context-Aware Predictive Optimization of Non-Functional Requirement Conflicts in DevOps Using Machine Learning
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 132938-132950
Open AccessContext-Aware Predictive Optimization of Non-Functional Requirement Conflicts in DevOps Using Machine Learning
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- p. 1
A Probabilistic Framework for Early-Stage Scalability Analysis in the Software Development Life Cycle
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- p. 1
Open AccessA Probabilistic Framework for Early-Stage Scalability Analysis in the Software Development Life Cycle
- 27 August 2025
- Published by Springer Nature in Innovations in Systems and Software Engineering
- p. 1-10
Pseudo CNN: An Updated CNN Based Semi Supervised Learning Model for Leaf Disease Detection
- 27 August 2025
- Published by Springer Nature in Innovations in Systems and Software Engineering
- p. 1-10
Pseudo CNN: An Updated CNN Based Semi Supervised Learning Model for Leaf Disease Detection
- 24 August 2025
- Published by Springer Nature
- p. 290-306
Optimizing Kidney Stone Detection: Fusion of CNN and SVM for Enhanced Performance in Medical Imaging
- 24 August 2025
- Published by Springer Nature
- p. 290-306
Optimizing Kidney Stone Detection: Fusion of CNN and SVM for Enhanced Performance in Medical Imaging
- Open Access
- 25 July 2025
- Published by College of Science for Women, University of Baghdad in Baghdad Science Journal
- Vol. 22 (7) , 2438-2457
A Study on Machine Learning Based Parkinson Disease Prediction
Parkinson's Disease, a neurodegenerative disorder, is one of the major chronic health issues in the world. It causes a severe disorder that affects majorly muscle control but it can also be the reason of affecting senses, cognitive ability and cognitive health. Approximately, 90% of the Parkinson affected people face speech difficulty. Conventional diagnosis methods may be biased which may result in wrong diagnosis of Parkinson's Disease because symptoms are usually elusive. This study aims to assess how well different machine learning algorithms can predict Parkinson's disease using vowel phonation data, with the goal of early detection and more accurate patient assessments. Different machine learning algorithms, like Random Forest, Logistic Regression, Decision Tree, Support Vector Machine, and Boosting algorithms (Gradient, Extreme Gradient, Light Gradient, and Categorical), are evaluated for their prediction ability. Based on vowel phonation data, Random Forest achieved the highest accuracy of 98.4% among the evaluated classifiers in predicting Parkinson's disease. It highlights the prominence of machine learning application for early detection of Parkinson's disease accurately. This research helps create a better way to assess patients' risk of Parkinson's disease, leading to a clearer understanding and supporting future studies in this area.
- 25 July 2025
- Published by College of Science for Women, University of Baghdad in Baghdad Science Journal
- Vol. 22 (7) , 2438-2457
Open AccessA Study on Machine Learning Based Parkinson Disease Prediction
Parkinson's Disease, a neurodegenerative disorder, is one of the major chronic health issues in the world. It causes a severe disorder that affects majorly muscle control but it can also be the reason of affecting senses, cognitive ability and cognitive health. Approximately, 90% of the Parkinson affected people face speech difficulty. Conventional diagnosis methods may be biased which may result in wrong diagnosis of Parkinson's Disease because symptoms are usually elusive. This study aims to assess how well different machine learning algorithms can predict Parkinson's disease using vowel phonation data, with the goal of early detection and more accurate patient assessments. Different machine learning algorithms, like Random Forest, Logistic Regression, Decision Tree, Support Vector Machine, and Boosting algorithms (Gradient, Extreme Gradient, Light Gradient, and Categorical), are evaluated for their prediction ability. Based on vowel phonation data, Random Forest achieved the highest accuracy of 98.4% among the evaluated classifiers in predicting Parkinson's disease. It highlights the prominence of machine learning application for early detection of Parkinson's disease accurately. This research helps create a better way to assess patients' risk of Parkinson's disease, leading to a clearer understanding and supporting future studies in this area.
- 1 October 2024
- Published by Springer Nature
- p. 339-360
A Study of Heart Disease Prediction Using Machine Learning
- 1 October 2024
- Published by Springer Nature
- p. 339-360
A Study of Heart Disease Prediction Using Machine Learning
- 18 January 2024
- Published by Springer Nature
- p. 227-242
Deep Learning-Based Approach for Plant Disease Classification
- 18 January 2024
- Published by Springer Nature
- p. 227-242
Deep Learning-Based Approach for Plant Disease Classification