Institution profile

Central University of Rajasthan

Academic institutionasia · in
Official website
Research library3linked papers
Opportunities0open roles
Selected work

Representative Papers

Assessing Reliability of BERT-Based Models on Question Answering Tasks

Aug 11, 2026

This study addresses the lack of systematic evaluation of reliability in question-answering models by proposing a novel framework that integrates Monte Carlo Dropout (MCD) with input semantic paraphrasing perturbations, requiring no modification to the standard inference pipeline. The authors conduct a comprehensive analysis of BERT, RoBERTa, ALBERT, and DistilBERT on the SQuAD and QuAC datasets. Experimental results demonstrate that MCD effectively captures prediction stability, with RoBERTa exhibiting the highest reliability, while ALBERT and DistilBERT show notably lower stability. This work not only validates MCD as a robust metric for model reliability but also establishes a new paradigm for evaluating the robustness of question-answering systems.

0 citationsRead paper

CVE-TTP KG: Knowledge Graph Linking Software Vulnerabilities to Attack Behaviors

Jun 30, 2026

This study addresses the critical gap in existing vulnerability databases, which lack effective linkage to adversary tactics and techniques, thereby limiting deep understanding of threat behaviors. To bridge this gap, the work presents the first systematic construction of a CVE–TTP knowledge graph by semantically associating Common Vulnerabilities and Exposures (CVEs) with MITRE ATT&CK tactics and techniques. This is achieved through an integrated approach combining CySecBERT and other Transformer-based models with a pipeline-style, span-based joint extraction framework. The authors release a large-scale annotated dataset, achieving macro F1 scores of 96.16% and 87.71% for tactic and technique identification, respectively, and 0.86 and 0.99 for entity and relation extraction. The resulting knowledge graph comprises 24,820 entities and 43,608 relations, and is structurally visualized using Neo4j.

0 citationsRead paper
Recent publications

Latest Papers

Assessing Reliability of BERT-Based Models on Question Answering Tasks

Aug 11, 2026

This study addresses the lack of systematic evaluation of reliability in question-answering models by proposing a novel framework that integrates Monte Carlo Dropout (MCD) with input semantic paraphrasing perturbations, requiring no modification to the standard inference pipeline. The authors conduct a comprehensive analysis of BERT, RoBERTa, ALBERT, and DistilBERT on the SQuAD and QuAC datasets. Experimental results demonstrate that MCD effectively captures prediction stability, with RoBERTa exhibiting the highest reliability, while ALBERT and DistilBERT show notably lower stability. This work not only validates MCD as a robust metric for model reliability but also establishes a new paradigm for evaluating the robustness of question-answering systems.

0 citationsRead paper

CVE-TTP KG: Knowledge Graph Linking Software Vulnerabilities to Attack Behaviors

Jun 30, 2026

This study addresses the critical gap in existing vulnerability databases, which lack effective linkage to adversary tactics and techniques, thereby limiting deep understanding of threat behaviors. To bridge this gap, the work presents the first systematic construction of a CVE–TTP knowledge graph by semantically associating Common Vulnerabilities and Exposures (CVEs) with MITRE ATT&CK tactics and techniques. This is achieved through an integrated approach combining CySecBERT and other Transformer-based models with a pipeline-style, span-based joint extraction framework. The authors release a large-scale annotated dataset, achieving macro F1 scores of 96.16% and 87.71% for tactic and technique identification, respectively, and 0.86 and 0.99 for entity and relation extraction. The resulting knowledge graph comprises 24,820 entities and 43,608 relations, and is structurally visualized using Neo4j.

0 citationsRead paper