Research on network public opinion resonance based on sentiment analysis
In order to cope with the complex and changing public opinion environment,the evolution of single public opinion event occurring in isolation into a cluster or a set of multiple public opinion events owing certain same characteristics was studied,and a resonance model of network public opinion for multiple public opinion events with the same topic and same emotion was established.Through crawling the Sina-Weibo data including 2022 Tangshan beating incident and the victim's first voice,as well as the torrential rain in Henan in 2021 and the torrential rain in Hebei in 2023,the BosonNLP sentiment analysis of the comment data was conducted to obtain its sentiment score.By utilizing the sentiment score as the model param-eter,two different types of cases were tested respectively.The results show that the primary public opinion events and seconda-ry public opinion events work together to cause the netizens'emotional infection and conflict,thus the network public opinion resonance occurs,and the heat generated by the resonance is much higher than that of single event.The netizen's attitude value,as well as the relevance of issue index,and the value of the event's heat,are the keys to judge whether the resonance occurs or not.The research results can enrich the theories of network public opinion and social physics,and provide a refer-ence for constructing a research framework of public opinion resonance.
network public opinionsentiment analysisstochastic resonanceLangevin equation
万方数据
维普
