University of Cambridge > Talks.cam > CUED Speech Group Seminars > Natural Language Interpretation in Dialogue Systems

Natural Language Interpretation in Dialogue Systems

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

If you have a question about this talk, please contact Dr Kate Knill.

Seminar on zoom

Abstract: Task-oriented dialogue systems are natural language interfaces that help the user to achieve a task, such as finding information from a data source, making a food order, or solving a problem. Natural language, a universal communication interface, is the input to a dialogue system. The flexibility of natural language, which allows ambiguity and multiple correct ways to represent the same information, presents an interpretation challenge to such systems. In a referring expression, a user may identify a previously discussed entity by specifying its properties, e.g. `the Italian place’ or `the restaurant in the centre’. An intelligent system should correctly handle a user’s references to the previously mentioned entities.

In this talk, I will first give an overview of the approaches used to design a dialogue manager that relies on natural language understanding for interpreting user input. Next, I will describe the novel action state update approach (ASU) for utterance interpretation. Using knowledge of the domain, ASU interprets a user utterance without a domain-specific natural language understanding component. We train the interpretation model using active learning on simulated dialogues in the restaurant search domain. With both user-simulated and interactive human evaluations, we show that the ASU approach successfully interprets user utterances in a dialogue system, including those with referring expressions.

Bio: Dr. Svetlana Stoyanchev is a Research Engineer at the Speech Technology Group in Cambridge Research Lab, Toshiba Europe. She received her Ph.D. in Computer Science from Stony Brook University. She held research positions at The Open University, Columbia University, AT&T Labs Research, and Interactions Corporation. She served as a member of IEEE Speech and Language Processing Technical Committee and the program committees for research conferences and workshops, including SIG Dial, ACL , AAAI, COLING , and Interspeech. Her research interests include conversational and multimodal interfaces, natural language understanding and generation, argumentation in dialogue, and error recovery in human-computer communication.

This talk is part of the CUED Speech Group Seminars series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2021 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity