Hot Spot Information Auto-detection Method of Network Public Opinion

According to the needs of network public opinion research in public safety, using Chinese word segmentation technology in the field of emergency management, this paper presents a hot spot information auto-detection method of network public opinion based on ICTCLAS, which inputs the text of the news corpus, uses Chinese word segmentation and word frequency statistics, gives the hot spot information list of through the removal of stop words and the merged of multiple keywords. It is verified through two emergency examples that the practicality and reliability of the method.