University of Cambridge > Talks.cam > NLIP Seminar Series > Interpretable Multi-hop Question Answering

Interpretable Multi-hop Question Answering

Add to your list(s) Download to your calendar using vCal

  • UserZhenyun Deng (University of Cambridge) World_link
  • ClockFriday 26 May 2023, 12:00-13:00
  • HouseComputer Lab, SS03.

If you have a question about this talk, please contact Michael Schlichtkrull.

Abstract:

Multi-hop question answering (QA) requires reasoning over multiple documents to answer a complex question and provide explanations for the answer. Most existing QA systems use an end-to-end model to answer natural language questions, making it difficult to demonstrate that a model has performed the desired reasoning to predict the correct answer. As a result, most QA systems have fallen short of the goal of answering multi-hop questions in an interpretable fashion using an acceptable reasoning pattern. My research attempts to tackle two distinct problems in multi-hop QA: i) how to build an interpretable QA system to answer a complex multi-hop question; ii) how to learn knowledge for the new QA tasks while retaining the knowledge learned on previous QA tasks. The goal of this research is to help humans better understand and trust the mechanism of a QA system using multi-hop reasoning.

Bio:

I’m Zhenyun Deng, a postdoctoral research associate working with Prof. Andreas Vlachos on automated fact checking at the Department of Computer Science and Technology at the University of Cambridge. Before that, I received my PhD from the University of Auckland, New Zealand. My PhD research focuses on interpretable multi-hop question answering, which aims to reason over multiple documents to answer a complex question and provide explanations for the answer.

This talk is part of the NLIP Seminar Series series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2024 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity