| Popis: |
With the advancement of computer technology, network opinions have gradually penetrated into every aspect of society. Network Forum as a major carrier of online public opinion communication, people often post their views on popular issues of social concern. However, the current public opinion in network forum does not factor in the time span of posts and the popularity of posters. In order to address the deficiencies of traditional post analysis, this paper studies the clustering algorithm based on co-word analysis technology and bisecting k-means, as well as the analysis of time span and popularity. Additionally, the current hotspots of the network are extracted and analyzed through experiments. The result of the propagation of public opinion shows an obvious peak over time, and the hot development curve drawn by this experiment is consistent with the facts, thereby validating it and proving its efficacy. |
