Building Trustworthy Intelligent Information Access Systems

3^rd March 2023, 13:00 Ashton Lecture Theatre
Dr. Procheta Sen
Department of Computer Science, University of Liverpool

Abstract

The talk will have two parts. In the first part the focus will be on developing Intelligent information access systems (IIAS). IIAS helps users to access information in an efficient way. IIAS aims to anticipate a user's inherent information need even before the user explicitly asks for it. IIAS tracks a user's activity to anticipate a user's information need in advance and then uses an information retrieval system to provide useful information resources to the user. The objective of IIAS is to reduce a user's effort and to increase user satisfaction.

The second part of the talk will be about explaining intelligent systems. With recent advancements of deep learning and machine learning techniques it has become important to explain an AI model to a user. More specifically, the focus will be on methodologies that can be used to investigate the output of retrieval models from the perspective of an information retrieval practitioner.

Biography

I am a Lecturer (Assistant Professor) in the Department of Computer Science at University of Liverpool (UOL), United Kingdom. I am a part of the Natural Language Processing group at UOL. Before joining UOL, I worked with Emine Yilmaz as a postdoctoral researcher in the Web Intelligence Group at University College London (UCL). I finished my PhD. in 2021 from Dublin City University, Ireland. My PhD. supervisor was Gareth J.F. Jones. I was affiliated with ADAPT centre Ireland during my PhD.

I broadly work in the area of Information Retrieval (IR), Natural Language Processing (NLP) and Machine Learning (ML). More specifically, I am interested in developing intelligent information access systems using IR and NLP techniques. I am also interested in making intelligent systems fair and explainable.