Addressing the Language Gap in Conversational Systems


Target: ₹ 40,00,000 ($ 50000)

Raised: ₹ 0 ($ 0)

No. of Donors : 0

Completion Date: 31st December 2026


Dr. Maunendra Sankar Desarkar, Associate Professor and Head of the AI Department, focuses on addressing language gaps in conversational systems. The project aims to leverage Large Language Models (LLMs) for domain-specific applications, particularly in regional languages, addressing the limitations of rule-based systems.


The project aims to harness Large Language Models (LLMs) for domain-specific conversational systems, emphasizing regional languages. It seeks to develop goal-oriented systems, covering tasks like intent identification, dialog state tracking, and dialog generation. With a focus on politeness, empathy, and informativeness, the project aims to benchmark LLMs' potential for diverse languages. The current work includes NLP for low-resource languages, and past achievements include Natural Language Generation for low-resource languages and a Transport-domain chatbot for Hindi-English queries, both at different Technology Readiness Levels (TRL).

Timeline & Budget:

Year 1: Project personnel (PhD, MTech) - INR 7 Lakhs, Equipment (Workstations) - INR 10 Lakhs, Travel and conference - INR 1 Lakh, Contingency - INR 1 Lakh (Total: INR 19 Lakhs)

Year 2: Project personnel - INR 7 Lakhs

Year 3: Project personnel - INR 7 Lakhs

Total Budget: INR 40 Lakhs

Proposer Details:

Dr. Maunendra Sankar Desarkar, Associate Professor, Department of CSE, and Head of the Department, Department of AI.