Clustering of Short Text in Micro-blog Based on K-means Algorithm

Based on K-means algorithm, this paper proposed a short text clustering method. First of all, data of short texts on the Internet are collected by using the web crawler. Then, they are preprocessed, for example, irrelevant contents like noisy data, punctuation and stop words, are removed. After that, word segmentation is carried out on the preprocessed short texts, and distributed expression is carried out on the segmented words. Finally, these texts are clustered and sorted on the basis of K-means algorithm. According to the experiment results, methods put forward in the paper are appropriate for short text clustering.