Clustering Efficient Method on Mass Chinese Text Based on Semantic Concept

In the current thinking of the Chinese text clustering, most clustering algorithms are limited by the data´s scalability and the results´ interpretability. This paper presents an efficient Chinese text clustering method based on semantic concepts. This method, proceeding from the text itself, by using classified hierarchy Subject Word in Thesaurus of Modern Chinese, extracts the conceptional tuple from a high-dimensional text vector collection to form the high-level concept expressing clustering results. Then samples are divided based on these high-level concepts which indicates the entire text clustering process has completed. On the premise of ensuring the clustering results´ accuracy, this method can greatly reduce the number of data needing to be processed and improve the clustering algorithms´ scalability. The experimental results show that this clustering algorithm has achieved a satisfactory clustering result and a higher implementation efficiency as well.