Public Opinion Analysis of Emergency on Weibo Based on Improved CSIM: The Case of Tianjin Port Explosion

Nowadays, Weibo, the most popular and commonly used microblog, has already played an important role of a public opinion field in China. Especially when emergency occurs, a public opinion storm will be raised on Weibo. An improved CSIM algorithm was put forward to help analyze the public opinions by clustering the messages. The Tianjin Port Explosion was chosen as an example and the related original posts and hot comments were collected as corpus. By analyzing the clustering result, the hot topics were identified and evolution patterns of public opinions on emergency were explored. According to the clustering result, the public first concerned about “the description of the explosion”, “mourning and prayer” and “the discussion about the responsibility of media”. However, in terms of quantity, the public most concerned about “the evaluation of government measures”. The next was “mourning and prayer”, while the third was “the description of the explosion”.