基于主题子空间的文本模糊C均值聚类方法

Fuzzy C-means Text Clustering Based on Topic Sub-space

模糊C均值聚类作为聚类的一种有效方法在数据挖掘和信息检索等领域得到广泛的应用,初始中心和初始隶属度矩阵的建立是决定模糊C均值聚类效果的关键.本文提出一种基于文本主题空间的模糊C均值聚类算法TS2FCM(Topic Sub-Space based Fuzzy C-Means),通过对能够代表文本主题的关键短语(salient phrase)的提取来建立主题子空间,利用主题子空间中的文本向量来提取初始中心和初始隶属度矩阵.实验表明,TS2FCM取得了较好的聚类效果.

As an effective method of clustering, Fuzzy C-Means（FCM） clustering is applied in many areas such as data mining and information retrieval. The key of FCM effect is the establishment of initial centers and subordination degree matrix. This paper proposes a TS2FCM（Topic Sub-Space based Fuzzy C-Means） algorithm which extracts salient phrases to build topic sub- space and then extract initial centers and subordination degree matrix. The experiment indicates that TS2FCM obtains good clustering effect.