The System of the Dissemination Characteristics of Internet Public Opinion Big Data Based on Artificial Intelligence

In the era we live in today, the network is often used to analyze a large number of complex systems. With the development of the information society, there are more and more ways to disseminate public information through social networks. Public opinion dissemination refers to the process of disseminating public opinion information through social networks. Because the dissemination of public opinion is the basis for the exchange of ideas among multiple communicators of public opinion, the network community will certainly have an impact on the dissemination and development of public opinion. This article is based on artificial intelligence to study the network public opinion big data dissemination characteristic analysis system, introduces the network public opinion analysis system based on the characteristics of the network public opinion, introduces in detail multiple methods and clustering algorithms for extracting the text information of Internet public opinion, and proposes the Kmeans + Canopy + semantic similarity algorithm, and uses the A event to compare the parameters of the network clustering coefficient, the correlation measure and the degree centrality measure, and the performance of the Kmeans + Canopy algorithm and the Kmeans + Canopy + semantic similarity algorithm. The results of the experiment found that the clustering coefficient of “People’s Daily” in the network dissemination of A event was 0.038, which was the highest among all nodes. It shows that 3.8% of the nodes established by the “People’s Daily” can interact one-to-one to deliver information and intelligence resources. Although the complexity of the algorithm has increased and the time consumed by the system has increased, the accuracy of clustering has been improved, especially for cultural articles, the accuracy rate has been as high as 75%, and entertainment articles can reach up to 70%, and stabilize at around 70%.