Microblog emotion analysis method combining microblog weak relationship

Abstract

The invention discloses a microblog emotion analysis method combining microblog weak relationship, and belongs to the technical field of text classification. The method comprises the following steps:step (1): calculating an emotion consistency relationship matrix between microblogs, and establishing a microblog emotion analysis model based on the emotion consistency; step (2): calculating a microblog emotional infectivity matrix, and establish a microblog emotion analysis model based on the emotional infectivity; step (3): using a community partitioning algorithm to calculate the microblog weak relationship matrix, and establishing a microblog emotion analysis model based on the microblog weak relationship; step (4): combining the steps (1), (2) and (3) to obtain a comprehensive model, and then adding the emotion analysis model based on the microblog text content and performing regularization to obtain a final model; step (5): optimizing the final model and solving a microblog emotionanalyzer. The microblog emotion analysis method combining microblog weak relationship in the invention not only utilizes the microblog matrix, but also considers the text content of the microblog, which is more accurate and credible.

Description

technical field [0001] The invention belongs to the technical field of text classification, and in particular relates to a microblog sentiment analysis method combined with microblog weak relationships. Background technique [0002] According to the survey of "2010 China Internet Public Opinion Analysis Report", the number of replies on Sina Weibo reached 2.6 million on the topic of "Tencent and 360 War" among the hot spots on the Internet in 2010, and the number of related comments on "Shanghai World Expo" also reached 1.06 million articles, far exceeding the number of news reports, communities, and forums. On July 23, 2011, the Wenzhou high-speed train incident received 2.82 million comments on Sina Weibo, and 6.84 million replies on Tencent Weibo. It can be seen that Weibo has become one of the important media for the dissemination of public opinion. Sentiment analysis on Weibo is the main means to grasp the emotional feedback of netizens on emergencies. [0003] With t...

AI Technical Summary

Problems solved by technology

However, this method has poor operability

Moreover, in the Internet age, the emotional attribution of emoticons is difficult to divide or changes rapidly, which reduces the credibility of sentiment analysis results.

Images