AUTHOR=Zhang Yang , Lian Ji-Qing , Li Ren-De , Duan Hong-Tao 

TITLE=Research on the evolution of netizens’ comment focus in university online public opinion: KTF-BTM topic model with topic-temporal-focus framework

JOURNAL=Frontiers in Physics

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/physics/articles/10.3389/fphy.2023.1251386

DOI=10.3389/fphy.2023.1251386

ISSN=2296-424X

ABSTRACT=Nowadays, Study of comments in MicroBlog online public opinion is of great significance for relevant departments in managing public opinion, due to the increasing influence of online public opinion on the Internet. To study the evolutionary characteristics of netizens' comment focus in university online public opinion, this paper proposes a three-stage framework called Topic-Temporal-Focus . In the topic mining stage, this paper utilizes the KTF-BTM model to improve the accuracy of topic recognition, which effectively enhances the quality of analysis. In the temporal segmentation stage, time periods are divided at intervals of 4 hours, and the identified topics are matched with each comment text to create a topic-temporal list. Finally, in the focus recognition stage, the content and evolution patterns of netizens' comment focus within shorter time sequences are explored by analyzing the data characteristics of the topic-temporal list. The KTF-BTM model proposed in this paper significantly improves the quality of topic recognition for short texts. The three-stage framework of Topic-Temporal-Focus overcomes the problem of sparse comment text data within shorter time periods and successfully classifies the topic classification problem within short time sequences. By combining these approaches, this paper effectively mines the evolutionary characteristics of netizens' comment focus in university online public opinion.