Share Email Print
cover

Proceedings Paper

Ontology-based topic clustering for online discussion data
Author(s): Yongheng Wang; Kening Cao; Xiaoming Zhang
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

With the rapid development of online communities, mining and extracting quality knowledge from online discussions becomes very important for the industrial and marketing sector, as well as for e-commerce applications and government. Most of the existing techniques model a discussion as a social network of users represented by a user-based graph without considering the content of the discussion. In this paper we propose a new multilayered mode to analysis online discussions. The user-based and message-based representation is combined in this model. A novel frequent concept sets based clustering method is used to cluster the original online discussion network into topic space. Domain ontology is used to improve the clustering accuracy. Parallel methods are also used to make the algorithms scalable to very large data sets. Our experimental study shows that the model and algorithms are effective when analyzing large scale online discussion data.

Paper Details

Date Published: 14 March 2013
PDF: 5 pages
Proc. SPIE 8768, International Conference on Graphic and Image Processing (ICGIP 2012), 87683T (14 March 2013); doi: 10.1117/12.2011097
Show Author Affiliations
Yongheng Wang, Hunan Univ. (China)
Kening Cao, Hunan Univ. (China)
Xiaoming Zhang, Hunan Univ. (China)


Published in SPIE Proceedings Vol. 8768:
International Conference on Graphic and Image Processing (ICGIP 2012)
Zeng Zhu, Editor(s)

© SPIE. Terms of Use
Back to Top