A Survey on Composite Social Networks Mining

PhD Qualifying Examination


Title: "A Survey on Composite Social Networks Mining"

by

Mr. Erheng Zhong


Abstract:

Social network analysis (SNA) has attracted many research interests in 
past years due to the rapid development of on-line social networks. Such 
analysis can help people understand the user behavior, network structures 
and information flow. It can also promote commercial applications, ranging 
from recommendation, on-line advertisement to social marketing. The major 
difference between social and traditional networks is that social networks 
are usually composite, where people may exist in multiple social networks. 
Such a property leads to two research issues. The first one is that each 
user in social networks may share different relationships with their 
neighbors, which require researchers to use multi-relational knowledge to 
perform comprehensive network analysis while taking into account their 
different contextual information. Secondly, the composite property sheds 
light on solving the ``sparsity'' problem in social networks, by 
considering the shared nodes among networks as the bridge and exploiting 
cross-network knowledge transfer. In this survey, we start from research 
on traditional social network analysis. We then review the recent research 
works on different composite social network mining tasks from 
multi-relational and cross-network aspects and organize the related 
literature into a structured presentation. Finally, we discuss some 
possible research issues.


Date:                   Wednesday, 30 November 2011

Time:                   10:00am - 12:00noon

Venue:                  Room 3405
                         lifts 17/18

Committee Members:	Prof. Qiang Yang (Supervisor)
                         Prof. Nevin Zhang (Chairperson)
 			Dr. Sunghun Kim
 			Dr. Raymond Wong


**** ALL are Welcome ****