Debasmita Bhattacharya

  • Scholarship in 2022 at Columbia University

About Debasmita Bhattacharya’s work

Debasmita Bhattacharya conducts research at the intersection of natural language processing (NLP) with linguistics and social science, with a focus on the social and psycholinguistic factors that drive code-switching between different languages and dialects of English.

By developing multilingual learning systems that are robust to code-switched linguistic data, Bhattacharya seeks to facilitate cross-lingual and cross-cultural communication and understanding in an increasingly globalised world. Solving problems in multilingual learning is a step towards more robust multilingual models, and greater inclusivity in NLP, where the current state-of-the-art work focuses almost entirely on standard dialects of English.

Bhattacharya is also interested in researching whether NLP methods are fair, especially with respect to bias that relates to demographic identity in models and datasets. She plans to investigate how to extend existing methods towards interdisciplinary computing for social good. This could, for example, help make social media platforms safer and fairer for millions of users.

Biography

Debasmita Bhattacharya is a PhD scholar in Columbia University’s Spoken Language Processing Group. Debasmita earned her BSc and MEng at Cornell University, where she majored in computer science and minored in linguistics.