Paper accepted at EMNLP 2025: DIWALI – Diversity and Inclusivity aWAre cuLture specific Items for India

Congratulations to the authors Pramit Sahoo, Maharaj Brahma, and Dr. Maunendra Sankar Desarkar on this outstanding achievement! (Pramit Sahoo and Maharaj Brahma contributed equally to this work.)

We are delighted to announce that the paper titled “DIWALI - Diversity and Inclusivity aWAre cuLture specific Items for India: Dataset and Assessment of LLMs for Cultural Text Adaptation in Indian Context” has been accepted at EMNLP 2025 (Empirical Methods in Natural Language Processing), a CORE Rank A* conference.

Key Highlight:

This work addresses the lack of culturally grounded evaluation resources for Large Language Models (LLMs) by introducing DIWALI, a novel dataset comprising 8,817 culture-specific items collected from 36 subregions of India across 17 cultural facets. The study further evaluates LLMs on the cultural text adaptation task, highlighting key gaps in deeper-level cultural adaptation and emphasizing the importance of building culturally aligned NLP systems.

Heartfelt congratulations to all the authors for this impactful contribution to culturally aware natural language processing research!