Analysis and Research of Onformation Dissemination in Social Networks Based on Artificial Intelligence

Aiming at features of the public opinion that exists online in the social media context, such as “extensive participation, real-time, rapid dissemination, diversified forms of expression, obvious emotional tendency, anonymity, individualism and fragmentation”, this paper is based on artificial intelligence, takes Sina Weibo data as the data source, and uses web crawler technology. The content of microblog forwarding and comments with the headline “rat head and duck neck” was collected as experimental samples. Through text mining technology, the life cycle evolution route of negative online public opinion in institutions of higher education was analyzed, and word frequency analysis and emotion analysis were carried out, so as to propose the governance path. The research results show that “Double peaks” can be seen in the evolution of unfavorable online public opinion in colleges and universities. When analyzing the sentiment behind unfavorable online public perceptions at colleges and institutions, the emotional tendency of Internet users is greatly affected by authoritative explanations, showing a characteristic of high volatility, which has certain reference significance for guiding negative Public opinion surveys that are conducted online in institutions of higher education.