Research on User-Oriented Short Text Clustering

摘要

To solve the problem of weak semantic description ability of short text feature and feature sparse of short text, this paper proposes a user-orientedshort text clustering method. In the pre-process, this method uses semantic-independent word dictionary to identify and remove semantic-independent word, and then uses semantics class dictionary to normalize semantics. In addition, this paper proposes position same meta similarity calculation method based on k-gram and proceeds hierarchical clustering on short texts finally. Results of experiment show that thealgorithm proposed by this paper can improve the effect of short text clustering.