A three-phase approach to document clustering based on topic significance degree

作者：

Highlights：

• We propose a three phase document clustering approach based on the best topic model.

• We present a definition of significance degree of topics for determining the best number of topics.

• We made the experiments to show the effectiveness and efficiency of our approach.

摘要

•We propose a three phase document clustering approach based on the best topic model.•We present a definition of significance degree of topics for determining the best number of topics.•We made the experiments to show the effectiveness and efficiency of our approach.

论文关键词：Document clustering,Topic model,K-means,K-means++

论文评审过程：Available online 15 July 2014.

论文官网地址：https://doi.org/10.1016/j.eswa.2014.07.014