Deep learning

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering

Final Year Thesis Oral Defense

Title: "Deep learning"

by

ZHOU Yingqi

Abstract:

In our final year thesis, we propose an open-domain dialogue system 
which provides a convenient channel for personalised information search. 
In contrast to task-oriented dialogue systems which cover only one 
specific domain, open-domain dialogue systems cover a broad range of 
topics. Furthermore, the text analysis component of this dialogue system 
extracts features from each query and the classifier chooses the best 
response collected from all topic-specific models, which enables users 
to query through the simple user-interface, rather than spend time 
finding the best suitable topic-specific apps. Moreover, the stored chat 
history will be utilized to analyse the user's preferences, then 
personalise the returned results.


Date            : 24 April 2018 (Tuesday)

Time            : 16:00 - 17:00

Venue           : Room 2304 (via lifts 17/18), HKUST

Advisor         : Prof. KWOK James Tin-Yau

2nd Reader      : Dr. SONG Yangqiu