Inferring Public and Private Topics for Similar Events

Event detection, extraction, and tracking can help people to better understand the event that happened in the world. Previous research focuses on mining single event. In this paper, we propose a topic model to infer the public and private topic from a group of similar events. Aiming at the consisten...

Full description

Saved in:
Bibliographic Details
Published inSemantic Web and Web Science pp. 153 - 159
Main Authors Ma, Xiaoli, Xia, Huan, Li, Juanzi, Wen, Xubo
Format Reference Book Chapter
LanguageEnglish
Published United States Springer New York 2013
SeriesSpringer Proceedings in Complexity
Subjects
Business mathematics & systems
Data mining
Mathematical physics
News Streams
Private Subjects
Public Subject
Topic Model
Topic Space
Online AccessGet full text

Cover

More Information
Summary:Event detection, extraction, and tracking can help people to better understand the event that happened in the world. Previous research focuses on mining single event. In this paper, we propose a topic model to infer the public and private topic from a group of similar events. Aiming at the consistency and mapping of topics, this model discriminates public and private topics by using Bernoulli distribution to determine the source of words. Experiment on earthquake dataset shows that our proposed algorithm can induce the public and private topics acceptable by users.
ISBN:1489997431
9781489997432
1461468795
9781461468790
ISSN:2213-8684
2213-8692
DOI:10.1007/978-1-4614-6880-6_13