Data Science TT Faculty Candidate Presentation Snigdha Chaturvedi, Ph.D. | Title: Structured Approaches to Natural Language Understanding

Tuesday, February 05, 2019
11:00 am
Floor/Room #: 
320

Data Science TT Faculty Candidate
Snigdha Chaturvedi, Ph.D.
Assistant Professor, University of California, Santa Cruz
Fuller Labs, 320 | 11:00 am | February 5, 2019


Title: Structured Approaches to Natural Language Understanding

Abstract: Despite recent advancements in Natural Language Processing, computers today cannot understand text in the ways that humans can. My research aims at creating computational methods that not only read but also understand text. To accomplish this, I develop machine-learning methods that incorporate linguistic cues as well as the context in which they appear to understand language. In this talk, I will discuss two specific applications of language understanding that focus on comprehension of narratives: (i) Choosing correct endings to stories, and (ii) Automatically generating narratives. 

Automatic narrative comprehension is a fundamental challenge in Natural Language Understanding, and can enable computers to understand social norms, human behavior and commonsense by processing large corpora of such texts. In the first part of the talk, I present a model that attempts to understand a story on three semantic axes: (i) its sequence of events, (ii) its emotional trajectory, and (iii) its plot consistency. We judge the model’s understanding by inquiring if, like humans, it can develop an expectation of what will happen next in a story and predict the correct ending from possible alternatives.

In the next part of the talk I will discuss the problem of interactive story generation. Interactive story-telling can find applications in real-world domains like entertainment and education. In this ongoing work we propose a deep learning model for generating a story sentence-by-sentence while the user provides ‘cue words’ that she wants to see in the next sentence. Our structured model learns to assimilate the previous sentence as well as the cue words for generating the next sentence. We evaluate the model by judging its capability to output coherent stories what satisfy user needs.

Apart from narrative understanding, I will also briefly discuss my ongoing and future work on applications of language understanding in domains like education, digital humanities and mental health care.

Bio: Snigdha Chaturvedi is an Assistant Professor in the Department of Computer Science and Engineering at the University of California, Santa Cruz. She specializes in the field of Natural Language Processing with an emphasis on developing methods for natural language understanding. Her research has been recognized with the IBM Ph.D. Fellowship (twice), a best paper award at NAACL, and first prize at ACM student research competition held at Grace Hopper Conference. Previously, she was a postdoctoral fellow at University of Illinois, Urbana Champaign, and University of Pennsylvania working with Professor Dan Roth. She earned her Ph.D. in Computer Science at University of Maryland, College Park in 2016 (advisor: Dr. Hal Daume III) and Bachelors of Technology from Indian Institute of Technology, Kanpur in 2009. She was also a Blue Scholar at IBM Research, India from 2009 to 2011.