Home // ICIW 2012, The Seventh International Conference on Internet and Web Applications and Services // View article

A Sub-topic Partition Method based on Event Network

Authors:
Wei Liu
Dong Wang
Wenjie Xu
Xujie Zhang
Zongtian Liu

Keywords: TDT,topic tracking,event network,community discovery,topic model,EN-MST

Abstract:
In TDT (Topic Detection and Tracking) of news story, a topic usually contains lots of events. Because of small granularity of event, the relationship between events and topic is not closely enough to be distinguished. In this paper, topic and news stories are described by using event networks, and a network clustering algorithm EN-MST based on minimum spanning tree is proposed to discover event communities in the network. Each community can be considered to be a sub-topic which represents an aspect of the large topic. As a bridge between events and topic, similarity calculation in the sub-topic level will improve the accuracy of topic detection and tracking.

Pages: 194 to 199

Copyright: Copyright (c) IARIA, 2012

Publication date: May 27, 2012

Published in: conference

ISSN: 2308-3972

ISBN: 978-1-61208-200-4

Location: Stuttgart, Germany

Dates: from May 27, 2012 to June 1, 2012