Sudip Kumar Naskar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Sudip Kumar Naskar has a total of 54 co-authors. Here are the top co-authors by number of shared publications:
- Sohom Ghosh
Jadavpur University, India
15 shared publications - Madhusudan Ghosh
Indian Association for the Cultivation of Science, Kolkata, India
8 shared publications - Partha Basuchowdhuri
Indian Association for the Cultivation of Science
8 shared publications - Debasis Ganguly
University of Glasgow, Glasgow, United Kingdom
7 shared publications - Asif Ekbal
School of AI and Data Science Indian Institute of Technology Jodhpur India
3 shared publications - Payel Santra
Indian Association for the Cultivation of Science Kolkata India
3 shared publications
Latest Publications
- 12 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 4th, 1-5
Exploring Zero-Shot Text Simplification for Bengali Science Text using Large Language Models
The increasing presence of complex scientific concepts and terminologies in Bengali school textbooks presents significant challenges for students, especially those from low-resource learning environments. In this paper, we explore automatic text simplification as a natural progression toward making these texts more accessible. We investigate the potential of Large Language Models (LLMs) to simplify scientific definitions in Bengali without requiring additional training data, adopting a zero-shot evaluation framework. Four advanced LLMs (GPT-5, Google Gemini, mT5_multilingual_XLSum, and mBART-large-50) are employed in this work. The study benchmarks their simplification performance across multiple criteria, including lexical reduction, syntactic simplification, semantic adequacy, and preservation of scientific meaning. Experimental results demonstrate that while transformer-based multilingual models like mT5 and mBART exhibit consistent structural simplification, GPT-5 and Gemini offer higher fluency and semantic fidelity. This work highlights the feasibility of using zero-shot LLMs for text simplification in low-resource educational settings and provides a comparative framework to guide future research in Bengali text simplification.
- 12 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 4th, 1-5
Exploring Zero-Shot Text Simplification for Bengali Science Text using Large Language Models
The increasing presence of complex scientific concepts and terminologies in Bengali school textbooks presents significant challenges for students, especially those from low-resource learning environments. In this paper, we explore automatic text simplification as a natural progression toward making these texts more accessible. We investigate the potential of Large Language Models (LLMs) to simplify scientific definitions in Bengali without requiring additional training data, adopting a zero-shot evaluation framework. Four advanced LLMs (GPT-5, Google Gemini, mT5_multilingual_XLSum, and mBART-large-50) are employed in this work. The study benchmarks their simplification performance across multiple criteria, including lexical reduction, syntactic simplification, semantic adequacy, and preservation of scientific meaning. Experimental results demonstrate that while transformer-based multilingual models like mT5 and mBART exhibit consistent structural simplification, GPT-5 and Gemini offer higher fluency and semantic fidelity. This work highlights the feasibility of using zero-shot LLMs for text simplification in low-resource educational settings and provides a comparative framework to guide future research in Bengali text simplification.
- Open Access
- 10 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 5202-5207
HF-RAG: Hierarchical Fusion-based RAG with Multiple Sources and Rankers
- 10 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 5202-5207
Open AccessHF-RAG: Hierarchical Fusion-based RAG with Multiple Sources and Rankers
- 19 April 2025
- Published by Springer Nature
- p. 250-262
RExAS: Relation Extraction Using Adaptive Self-attention
- 19 April 2025
- Published by Springer Nature
- p. 250-262
RExAS: Relation Extraction Using Adaptive Self-attention
- Open Access
- 18 April 2025
- Published by MDPI AG in Preprints.org
BASIR: Budget-Assisted Sectoral Impact Ranking — A Dataset for Sector Identification and Performance Prediction Using Language Models
Government fiscal policies, particularly annual union budgets, exert significant influence on financial markets. However, real-time analysis of budgetary impacts on sector-specific equity performance remains methodologically challenging and largely unexplored. This study proposes a framework to systematically identify and rank sectors poised to benefit from India’s Union Budget announcements. The framework addresses two core tasks: (1) multi-label classification of excerpts from budget transcripts into 81 predefined economic sectors, and (2) performance ranking of these sectors. Leveraging a comprehensive corpus of Indian Union Budget transcripts from 1947 to 2025, we introduce BASIR (Budget-Assisted Sectoral Impact Ranking), an annotated dataset mapping excerpts from budgetary transcripts to sectoral impacts. Our architecture incorporates finetuned embeddings for sector identification, coupled with language models that rank sectors based on their predicted performances. Our results demonstrate 0.605 F1-score in sector classification, and 0.997 NDCG score in predicting ranks of sectors based on post-budget performances. The methodology enables investors and policymakers to quantify fiscal policy impacts through structured, data-driven insights, addressing critical gaps in manual analysis. The annotated dataset has been released under CC-BY-NC-SA-4.0 licence to advance computational economics research.
- 18 April 2025
- Published by MDPI AG in Preprints.org
Open AccessBASIR: Budget-Assisted Sectoral Impact Ranking — A Dataset for Sector Identification and Performance Prediction Using Language Models
Government fiscal policies, particularly annual union budgets, exert significant influence on financial markets. However, real-time analysis of budgetary impacts on sector-specific equity performance remains methodologically challenging and largely unexplored. This study proposes a framework to systematically identify and rank sectors poised to benefit from India’s Union Budget announcements. The framework addresses two core tasks: (1) multi-label classification of excerpts from budget transcripts into 81 predefined economic sectors, and (2) performance ranking of these sectors. Leveraging a comprehensive corpus of Indian Union Budget transcripts from 1947 to 2025, we introduce BASIR (Budget-Assisted Sectoral Impact Ranking), an annotated dataset mapping excerpts from budgetary transcripts to sectoral impacts. Our architecture incorporates finetuned embeddings for sector identification, coupled with language models that rank sectors based on their predicted performances. Our results demonstrate 0.605 F1-score in sector classification, and 0.997 NDCG score in predicting ranks of sectors based on post-budget performances. The methodology enables investors and policymakers to quantify fiscal policy impacts through structured, data-driven insights, addressing critical gaps in manual analysis. The annotated dataset has been released under CC-BY-NC-SA-4.0 licence to advance computational economics research.
- Open Access
MiMIC: Multi-Modal Indian Earnings Calls Dataset to Predict Stock Prices
Predicting stock market prices following corporate earnings calls remains a significant challenge for investors and researchers alike, requiring innovative approaches that can process diverse information sources. This study investigates the impact of corporate earnings calls on stock prices by introducing a multi-modal predictive model. We leverage textual data from earnings call transcripts, along with images and tables from accompanying presentations, to forecast stock price movements on the trading day immediately following these calls. To facilitate this research, we developed the MiMIC (Multi-Modal Indian Earnings Calls) dataset, encompassing companies representing the Nifty 50, Nifty MidCap 50, and Nifty Small 50 indices. The dataset includes earnings call transcripts, presentations, fundamentals, technical indicators, and subsequent stock prices. We present a multimodal analytical framework that integrates quantitative variables with predictive signals derived from textual and visual modalities, thereby enabling a holistic approach to feature representation and analysis. This multi-modal approach demonstrates the potential for integrating diverse information sources to enhance financial forecasting accuracy. To promote further research in computational economics, we have made the MiMIC dataset publicly available under the CC-NC-SA-4.0 licence. Our work contributes to the growing body of literature on market reactions to corporate communications and highlights the efficacy of multi-modal machine learning techniques in financial analysis.
Open AccessMiMIC: Multi-Modal Indian Earnings Calls Dataset to Predict Stock Prices
Predicting stock market prices following corporate earnings calls remains a significant challenge for investors and researchers alike, requiring innovative approaches that can process diverse information sources. This study investigates the impact of corporate earnings calls on stock prices by introducing a multi-modal predictive model. We leverage textual data from earnings call transcripts, along with images and tables from accompanying presentations, to forecast stock price movements on the trading day immediately following these calls. To facilitate this research, we developed the MiMIC (Multi-Modal Indian Earnings Calls) dataset, encompassing companies representing the Nifty 50, Nifty MidCap 50, and Nifty Small 50 indices. The dataset includes earnings call transcripts, presentations, fundamentals, technical indicators, and subsequent stock prices. We present a multimodal analytical framework that integrates quantitative variables with predictive signals derived from textual and visual modalities, thereby enabling a holistic approach to feature representation and analysis. This multi-modal approach demonstrates the potential for integrating diverse information sources to enhance financial forecasting accuracy. To promote further research in computational economics, we have made the MiMIC dataset publicly available under the CC-NC-SA-4.0 licence. Our work contributes to the growing body of literature on market reactions to corporate communications and highlights the efficacy of multi-modal machine learning techniques in financial analysis.
- Open Access
- 16 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-13
Chronological Evaluation of Novel Methodology Extraction from AI Literature
- 16 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-13
Open AccessChronological Evaluation of Novel Methodology Extraction from AI Literature
- Open Access
AlpaPICO: Extraction of PICO Frames from Clinical Trial Documents Using LLMs
In recent years, there has been a surge in the publication of clinical trial reports, making it challenging to conduct systematic reviews. Automatically extracting Population, Intervention, Comparator, and Outcome (PICO) from clinical trial studies can alleviate the traditionally time-consuming process of manually scrutinizing systematic reviews. Existing approaches of PICO frame extraction involves supervised approach that relies on the existence of manually annotated data points in the form of BIO label tagging. Recent approaches, such as In-Context Learning (ICL), which has been shown to be effective for a number of downstream NLP tasks, require the use of labeled examples. In this work, we adopt ICL strategy by employing the pretrained knowledge of Large Language Models (LLMs), gathered during the pretraining phase of an LLM, to automatically extract the PICO-related terminologies from clinical trial documents in unsupervised set up to bypass the availability of large number of annotated data instances. Additionally, to showcase the highest effectiveness of LLM in oracle scenario where large number of annotated samples are available, we adopt the instruction tuning strategy by employing Low Rank Adaptation (LORA) to conduct the training of gigantic model in low resource environment for the PICO frame extraction task. Our empirical results show that our proposed ICL-based framework produces comparable results on all the version of EBM-NLP datasets and the proposed instruction tuned version of our framework produces state-of-the-art results on all the different EBM-NLP datasets. Our project is available at \url{https://github.com/shrimonmuke0202/AlpaPICO.git}.
Open AccessAlpaPICO: Extraction of PICO Frames from Clinical Trial Documents Using LLMs
In recent years, there has been a surge in the publication of clinical trial reports, making it challenging to conduct systematic reviews. Automatically extracting Population, Intervention, Comparator, and Outcome (PICO) from clinical trial studies can alleviate the traditionally time-consuming process of manually scrutinizing systematic reviews. Existing approaches of PICO frame extraction involves supervised approach that relies on the existence of manually annotated data points in the form of BIO label tagging. Recent approaches, such as In-Context Learning (ICL), which has been shown to be effective for a number of downstream NLP tasks, require the use of labeled examples. In this work, we adopt ICL strategy by employing the pretrained knowledge of Large Language Models (LLMs), gathered during the pretraining phase of an LLM, to automatically extract the PICO-related terminologies from clinical trial documents in unsupervised set up to bypass the availability of large number of annotated data instances. Additionally, to showcase the highest effectiveness of LLM in oracle scenario where large number of annotated samples are available, we adopt the instruction tuning strategy by employing Low Rank Adaptation (LORA) to conduct the training of gigantic model in low resource environment for the PICO frame extraction task. Our empirical results show that our proposed ICL-based framework produces comparable results on all the version of EBM-NLP datasets and the proposed instruction tuned version of our framework produces state-of-the-art results on all the different EBM-NLP datasets. Our project is available at \url{https://github.com/shrimonmuke0202/AlpaPICO.git}.
- 18 August 2024
- Published by Springer Nature
- p. 381-387
“The Absence of Evidence is Not the Evidence of Absence”: Fact Verification via Information Retrieval-Based In-Context Learning
- 18 August 2024
- Published by Springer Nature
- p. 381-387
“The Absence of Evidence is Not the Evidence of Absence”: Fact Verification via Information Retrieval-Based In-Context Learning
- Open Access
- 13 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 597-600
Generator-Guided Crowd Reaction Assessment
In the realm of social media, understanding and predicting post reach is a significant challenge. This paper presents a Crowd Reaction AssessMent (CReAM) task designed to estimate if a given social media post will receive more reaction than another, a particularly essential task for digital marketers and content writers. We introduce the Crowd Reaction Estimation Dataset (CRED), consisting of pairs of tweets from The White House with comparative measures of retweet count. The proposed Generator-Guided Estimation Approach (GGEA) leverages generative Large Language Models (LLMs), such as ChatGPT, FLAN-UL2, and Claude, to guide classification models for making better predictions. Our results reveal that a fine-tuned FLANG-RoBERTa model, utilizing a cross-encoder architecture with tweet content and responses generated by Claude, performs optimally. We further use a T5-based paraphraser to generate paraphrases of a given post and demonstrate GGEA's ability to predict which post will elicit the most reactions. We believe this novel application of LLMs provides a significant advancement in predicting social media post reach.
- 13 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 597-600
Open AccessGenerator-Guided Crowd Reaction Assessment
In the realm of social media, understanding and predicting post reach is a significant challenge. This paper presents a Crowd Reaction AssessMent (CReAM) task designed to estimate if a given social media post will receive more reaction than another, a particularly essential task for digital marketers and content writers. We introduce the Crowd Reaction Estimation Dataset (CRED), consisting of pairs of tweets from The White House with comparative measures of retweet count. The proposed Generator-Guided Estimation Approach (GGEA) leverages generative Large Language Models (LLMs), such as ChatGPT, FLAN-UL2, and Claude, to guide classification models for making better predictions. Our results reveal that a fine-tuned FLANG-RoBERTa model, utilizing a cross-encoder architecture with tweet content and responses generated by Claude, performs optimally. We further use a T5-based paraphraser to generate paraphrases of a given post and demonstrate GGEA's ability to predict which post will elicit the most reactions. We believe this novel application of LLMs provides a significant advancement in predicting social media post reach.