Dynamic topic detection and tracking based on knowledge base

From WikiPapers
Jump to: navigation, search

Dynamic topic detection and tracking based on knowledge base is a 2010 conference paper written in English by Wang S., Du J., Liang M., Chen L. and published in Proceedings - 2010 3rd IEEE International Conference on Broadband Network and Multimedia Technology, IC-BNMT2010.

[edit] Abstract

In order to solve the sparse initial information problem when the topic model was established ever before, this paper establishes the Wikipedia based news event knowledge base. Referring to this knowledge base, we calculate the weight of the news model, make the similarity measurement based on the time distance, make the clustering based on time line, and apply the dynamic threshold strategy to detect and track the topics automatically in the news materials. The experiment result verifies the validity of this method.

[edit] References

This section requires expansion. Please, help!

Cited by

Probably, this publication is cited by others, but there are no articles available for them in WikiPapers.