Document Clustering Method Using Weighted Semantic Features and Cluster Similarity

Abstract

In this paper, a document clustering method that use the weighted semantic features and cluster similarity is introduced to cluster meaningful topics from document set. The proposed method can improve the quality of document clustering because it can avoid clustering the documents whose similarities with topics are high but are meaningless between cluster and document by using the weighted semantic features. Besides, it uses cluster similarity to remove dissimilarity documents in clusters and avoid the biased inherent semantics of the documents to be reflected in clusters by NMF (non-negative matrix factorization). The experimental results demonstrate that the proposed method has better performance than other document clustering methods.