A Survey on Speaker Modeling in Neural Conversation Systems

PhD Qualifying Examination


Title: "A Survey on Speaker Modeling in Neural Conversation Systems"

by

Mr. Zhiliang TIAN


Abstract:

Human-to-machine conversation systems assign machines (chatbots) the 
intelligence of leading a conversation with humans. Most existing conversation 
systems respond to the user according to the user's historical utterances. Some 
researchers argue that conversation systems should consider the speaker's 
characteristics and status in the dialogues. The speaker's characteristic and 
status consists of the speaker's personality, speaking styles, and current 
emotional status. Such information helps conversation systems to generate 
appropriate and lively responses.

In this survey, we review the previous studies on neural conversation systems, 
which consider and model the information of the speakers. We first introduce 
the background, motivation, and challenges. Then a taxonomy is proposed based 
on the works we review. In a human-to-machine conversation scenario where users 
chat with the chatbots, the related research can be categorized into two 
subclasses: emotional conversation systems and personalized conversation 
systems. In the end, we will summarize new trends and potential future work to 
guide our research.


Date:			Tuesday, 15 December 2020

Time:                  	10:00am - 12:00noon

Zoom meeting:           https://hkust.zoom.com.cn/j/7491359443

Committee Members:	Prof. Nevin Zhang (Supervisor)
 			Dr. Brian Mak (Chairperson)
 			Prof. Fangzhen Lin
 			Dr. Yangqiu Song


**** ALL are Welcome ****