Social media user role classification method in public opinion event

Abstract

The invention discloses a social media user role classification method in a public opinion event, which comprises the following steps of: 1, acquiring user information in social media and extracting user characteristics, 2, designing a non-parametric Bayesian model and Dirichlet process hybrid model (h-DPMM) to model user roles, 3, designing Gibbs sampling inference parameters, analyzing the user roles of the social media in the public opinion event. The users can be classified into different roles by utilizing the content generated by the users in the online social media platform, so that the role types in a group can be comprehensively, quickly and accurately classified.

Description

technical field [0001] The invention relates to the technical field of role classification in online social media, in particular to a method for classifying roles of social media users in public opinion events. Background technique [0002] Online social media platforms have greatly enriched the way people generate, share and consume knowledge and information, and become the main medium for people to produce and disseminate information. Research shows that user-generated content is an important means of expressing inner qualities such as interests and attitudes. At the same time, user-generated content also shows the user's own living habits and role in the entire social networking platform. Classifying user roles is of great significance for classifying different types of users in social networks. [0003] When the existing technology classifies the roles of users in online media, most of them classify roles based on user behavior data, ignoring user-generated content, an...

AI Technical Summary

Problems solved by technology

[0003] When the existing technology classifies the roles of users in online media, most of them classify roles based on user behavior data, ignoring user-generated content, and failing to comprehensively utilize user-generated content and user behavior data, two important information reflecting user roles. , leading to inaccurate role classification; at the same time, the existing role classification methods require the number of predefined roles, but with the development of social networks, the number of predefined roles has brought great difficulties to role classification; in addition , most of the existing models regard each user as a single data point with the same feature structure, where the feature information of each user does not contain any user behavior or other attribute information, so it is not enough to analyze a large number of behaviors in social media data

Images