Study of topic mining for microblog based on CBMB-HL model
Huang Min, Ke Li · 2020
Because of the immediacy and interactivity of microblog, it became particularly important to analysis micro-blog theme. Based on the characteristic of the microblog texts, the relationship between microblog users and the context of comments, this paper proposes a new topic mining model-CBMB-HL model, which is based on multiple features and topic state association model, and can mine the content and comments of Weibo blog simultaneously. The experiments show that CBMB-HL model is more accurately and effectively than CBMB-LDA model to do the topic mining and can effectively discover the relationship between subjects.