********************************************************************


                                                     Seminar

             Department of Systems Engineering and Engineering Management
                                  The Chinese University of Hong Kong

------------------------------------------------------------------------------------------

 

 

 

Title

:

5SONGS-- A Song Recommendation System based on Personalized Song Taste Profile

 

 

 

Speaker

:

Prof. Yunqing Xia

 

 

Tsinghua University

 

 

 

 

 

Date

:

April 17th, 2008 (Thursday)

 

 

 

Time

:

2:00 p.m. -3:00 p.m.

 

 

 

Venue

:

Room 513

 

 

William M.W. Mong Engineering Building

 

 

(Engineering Building Complex Phase 2)

 

 

CUHK

 

 

 

------------------------------------------------------------------------------------------

Abstract:
 

Song recommendation seeks to actively suggest potential favorite songs to users. This research nowadays becomes a hot research topic due largely to the increasing demand of ubiquitous music access, especially via mobile phone. This talk presents a song recommendation system based on personalized song taste profile which is automatically learned from the song favorite list maintained by each user. To model user taste on song, two types of song attributes must be extracted, i.e. explicit attributes such as singer, composer and lyric writer and implicit attributes such as timber, rhythm, intensity and sentiment. The previous works indicate that audio-based methods are stronge in analysis on timber, rhythm and intensity while weak in sentiment analysis. It is thus proposed in this work to make use of textual information (i.e., song lyric) to detect song sentiment via machine learning approaches. However, problems are encountered when the word-based vector space model is adopted for sentiment analysis. The sentiment-based vector space model is therefore proposed in this talk to deal with the problems. Preliminary experiments show the new solution is very promising.


-------------------------------------------------------------------------------------------

Biography:
 

Prof. Yunqing Xia is associate professor in Tsinghua University, China. He obtained his PhD degree at Institute of Computing Technologies, Chinese Academy of Science, China in July 2001. He worked for University of Sheffield from January 2003 to September 2004 as postdoctoral researcher in Sheffield NLP group. From December 2005 to September 2006, he worked for The Chinese University of Hong Kong as postdoctoral fellow in Department of SE&EM. In October 2006, he joined Tsinghua University as associate professor within Center for Speech and Language Technologies (CSLT). His research interests cover natural language processing, computational linguistics and text mining. He is reviewer of some world class journals (e.g., ACM TALIP, IJCPOL, etc.) and distinguished international conferences (e.g., ACL, COLING, IJCNLP, etc.). He is standard member of IEEE computer science society, Association for Computational Linguistics (ACL) and Chinese Computer Federation. Currently, he is teaching the course for Natural Language Processing within Tsinghua University.


************************* ALL ARE WELCOME ************************

 

 

 

Host

:

Prof. Wong, Kam Fai

Tel

:

(852) 2609-332

Email

:

kfwong@se.cuhk.edu.hk

 

 

 

Enquiries

:

Prof. Nan Chen or Prof. Sean X. Zhou

 

:

Department of Systems Engineering and Engineering Management

 

 

CUHK

Website

:

http://www.se.cuhk.edu.hk/~seg5810

Email

:

seg5810@se.cuhk.edu.hk

 

 

 

********************************************************************