2024 to 2026
Human-in-the-Loop Federated Learning Framework for Sequence Labelling Task Targeted to Low-Resource Indian LanguagesSponsored by IIT Bhubaneswar
Research
I primarily explore Natural Language Processing, with a focus on Indian and other low-resource languages and on how large language models reason and plan.
Research areas
NLP for Indian and low-resource languages
Resources and models for Hindi, Bengali, Telugu and other Indian languages: WordNet enrichment, hypernymy detection, summarization, sentiment analysis, speech recognition and speech-to-speech translation.
Lexical semantics
Word senses, compositionality, co-hyponymy and lexical semantic change, using network and embedding methods.
Reasoning and planning with large language models
Benchmarks and methods for negation reasoning, common-sense reasoning, semantic role labelling and travel planning under constraints and disruptions.
Language technology for society
Legal language understanding, counterspeech generation, fake news detection, user engagement in news media, and privacy-preserving federated learning for sequence labelling.
Projects
Ongoing
2024 to 2027
Optimization of Algorithms for Voice Analysis System for Defence ApplicationsSponsored by DRDO
Completed
2023 to 2026
Interactive story/poetry writing and scene generation using Augmented RealitySponsored by VARCOE
2024 to 2025
Geospatial Knowledge Inference using LLMs for Enhanced Travel Decision-MakingSponsored by Microsoft
Prospective Ph.D. and MS (Research) applicants
See the department's broad research areas and admission page. You can also browse my publications to see the kind of work my group does.