An incremental clustering method of micro-blog topic detection

Micro-blog as a novel individual publication model over the internet, greatly promotes the open and interactive network information, but it has brought explosive growth on the information of micro-blog. Compared with the traditional topic, the text of micro-blog is shorter and less words, and the terms are not standard. Therefore the traditional topic detection method cannot work out effectively. In the paper, micro-blog account and network words are processed to reduce network garbage on micro-blog. Information of word, such as part of speech, frequency, and distribution, is used to extract feature of words. Incremental clustering model has been used to detect hot topic of micro-blog. The results show that the method improves the efficiency of the detection to a certain extent, and reduces the undetected rate and false detection rate. It can effectively discover hot topics on micro-blog in time.