A social network-oriented link prediction system and method

Abstract

The invention provides a link prediction system and method for a social network, and belongs to the field of data mining and social network analysis. On the basis of online users and a user friendly relational network, a social network link prediction model is constructed from the three aspects of user focused interests, information interaction and common adjacency users. The method comprises the steps that firstly, for multiple interest label characteristics of users in a social network, an LDA theme model is used for performing modeling on the users, and theme distribution for user behaviors is obtained; secondly, Gaussian weighting is used for modifying standard LDA, and the theme expression ability is improved; finally, by introducing a common adjacency user contribution algorithm defined by hidden naive Bayes, link prediction is performed on synthesis user behavior characteristics and network structure characteristics. The correlative dependence of the common adjacency users is more sufficiently considered, link prediction is performed on synthesis user behavior characteristics and network structure characteristics, and key factors built by links are found.

Description

Technical field [0001] The invention relates to the fields of data mining and social network analysis, and to user relationship analysis, in particular to a link prediction method oriented to social networks. Background technique [0002] With the continuous development of computer information technology and the rapid popularization of the Internet, social networks have attracted more and more people's participation and attention. In recent years, social networking sites have gradually become an important way for information dissemination and sharing. The relevant information left by users makes social networking a huge information platform, and the mining of these massive data has become a research hotspot. Among them, the research on user relationship analysis in social networks can help people better explain the evolution and discovery of network structure. [0003] At this stage, there are different aspects of research on user relationship analysis in social networks, includin...

AI Technical Summary

Problems solved by technology

In addition, the LDA model does not fully reflect the contribution of the network structure to link prediction. In fact, there is a certain dependency between common neighbors. Therefore, this dependency needs to be more fully considered to improve the accuracy of link prediction.

Images