- This event has passed.
{Seminar} @ CDS: #202 : 02nd June: “What Happens in India Stays in India!” – Fairness in NLP and ML in the Indian Context”
02 Jun @ 12:00 PM -- 1:00 PM
Department of Computational and Data Sciences
Department Seminar
SPEAKER : Shaily Bhatt (Google Research)
TITLE : “What Happens in India Stays in India!” – Fairness in NLP and ML in the Indian Context”
Date & Time : June 02, 2023, 12:00 Noon
Venue : # 202 CDS
===========================================================================================
ABSTRACT
Recent research has revealed undesirable biases in NLP and ML data and models. However, conventional algorithmic fairness is West-centric, as seen in its axes of disparities, values, and methods, and is not directly portable to other geo-cultural contexts. In this talk, I will focus on re-contextualizing fairness to the geo-cultural context of India. The first half will focus on why such a re-contextualization is necessary in the broader perspective of Algorithmic Fairness. Presenting an overview from Re-imagining Algorithmic Fairness in India and Beyond, I will touch upon the implicit assumptions of algorithmic fairness that do not hold true for the Indian context and provide a broad roadmap to re-contextualise data and models, empower oppressed communities, and enable Fair-ML ecosystems. In the second half, I will focus on our work on re-contextualizing NLP fairness in the context of India. I will touch upon the prominent axes of social disparities in India and how they manifest in language technologies and data. I will then dive deeper into building resources for NLP fairness evaluation in the Indian context and using them to demonstrate prediction biases along some of the axes. Finally, I will outline a holistic research agenda to re-contextualize NLP fairness research for the Indian context that accounts for societal, technological, and cultural aspects.
BIOGRAPHY
I am a Predoctoral Researcher working at the NLU Group at Google Research, India. My current research focuses on bias and fairness in multilingual models in the non-western context. Prior to this, I spent a year at Microsoft Research, India where I worked on interpretable evaluation and improvement of NLP models for multiple languages at scale. I graduated from BITS Pilani with a B.E in Computer Science in 2021. I am soon headed for my PhD in Fall 2023 at LTI, CMU.
Host Faculty: Dr. Danish Pruthi
==================================================================================
ALL ARE WELCOME