Public opinion popularity prediction method and system based on graph neural network

Abstract

The invention discloses a public opinion popularity prediction method and system based on a graph neural network, and the method comprises the following steps: firstly, defining popularity, and building a relation model between an information cascade graph and the popularity; obtaining cascade graph features based on the relation model cascade graph features; obtaining time sequence features based on the relation model cascade graph features; obtaining an emotion value of each time interval based on the time sequence features, and generating emotion sequence features; and connecting the cascade graph features, the time sequence features and the emotion sequence features, and calculating to obtain future popularity. The method can effectively predict the popularity of the public opinion event. The method is beneficial to the management of social events, reduces the economic loss, and solves the technical problems of low cascade graph representation efficiency and non-ideal prediction accuracy in the prior art.

Description

technical field [0001] The invention belongs to the technical field of network information security, and relates to a public opinion popularity prediction method and system based on a graph neural network. Background technique [0002] With the rapid development of the Internet, new media represented by online social platforms have become an important platform for the dissemination and fermentation of online public opinion, and the amount of information is also rapidly expanding at an unprecedented rate, which makes the spread of social events on the Internet Faster and more extensive. In order to better manage time in social networks and improve the governance level of network time, it is crucial to analyze the spread of events. Popularity prediction is a research focus in event propagation analysis. Effectively predicting the popularity of online content is very important for the governance and control of network events. [0003] Traditional popularity prediction methods...

AI Technical Summary

Problems solved by technology

In recent years, researchers have focused on using graph neural networks to learn cascaded graph features, however, cascaded graph representations in these graph neural network-based methods are usually generated by simple graph pooling strategies, which cannot distinguish the importance of users and lead to hierarchical Linked graphs represent inefficiencies

And there are few methods in the current research to add emotional factors, resulting in unsatisfactory prediction accuracy

Images