Evolution Analysis of Network Public Opinion Theme Based on LDA Model
Xiaoxiao Lu, AN Jun-xiu · 2022
Online public opinion is a collection of netizens' emotions, attitudes, opinions, opinions and so on. With the development of the Internet, the influence of online public opinion on social stability is increasing day by day. This paper takes the “COVID-19” event as an example, crawls the relevant news and comment data released by People's Daily, and firstly divides public opinion events into four stages according to the news popularity and life cycle theory: Tf-idf algorithm is used to strengthen the selection of key feature words in the corpus. Finally, LDA theme model is used to identify the topic of public opinion and mine the evolution law of network public opinion, which is helpful to effectively guide and control network public opinion and plays an important role in social stability.