Chinese microblog topic information processing method

Abstract

The invention discloses a Chinese microblog topic information processing method, and relates to reason analysis algorithms for emotional distribution of microblog events. The invention aims to solve the problems that a hierarchical clustering algorithm and a correction algorithm adopted in an existing microblog topic information processing method are low in accuracy and incapable of including event-related microblogs in correct topics. According to the Chinese microblog topic information processing method, event topics and related microblogs are mined with a hierarchical clustering ordering method of unsupervised learning and a microblog topic correction algorithm of semi-supervised learning, so that the purpose of performing emotional distribution statistics and analysis on the related microblogs is finally achieved. The Chinese microblog topic information processing method can perform microblog topic information processing more accurately. The present invention is applied to the microblog topic information processing field.The Chinese microblog topic information processing method is applied to the field of microblog topic information processing.

Description

technical field [0001] The invention relates to a method for processing microblog topic information. Background technique [0002] As an emerging social media platform, Weibo is also one of the most popular social media platforms in China. There are hundreds of millions of active users. More and more netizens choose to obtain and share information they are interested in on Weibo. In the face of big data of tens of millions per day on Weibo, it is very meaningful to analyze netizens’ opinions and attitudes towards a certain event. More and more scholars have begun to pay attention to the information behind such big data on Weibo. [0003] Since microblogging as a form of social media has not been in people's lives for a long time, there are not many related researches on the analysis of the causes of microblogging event emotion distribution at home and abroad. The current microblogging event mining methods mainly include, in 2011, Weng et al. used the relevant principle of w...

AI Technical Summary

Problems solved by technology

Compared with the rough work of the predecessors, the work of Abhimanyu and Anitha in 2014 (document [4] DasA, KannanA. In order to dig out hot topics in Twitter, by observing the commonality of Weibo events, three evaluation indicators are obtained, namely "diversity", "uniqueness" and "burstiness". , use the weakly labeled training corpus to fit the data distribution through a Gaussian mixture model, so as to output whether the candidate angle is a microblog event. Such a supervised learning topic extraction method can also achieve good results, but unfortunately this algorithm Clustering and sorting processing without topics involved

Images