Social network user recommendation method and system based on multiple dimensions

Abstract

The invention discloses a social network user recommendation method and system based on multiple dimensions. The method comprises the following steps: S1, extracting information of each user in a social network; s2, performing data preprocessing, and combining the preprocessed data into a text message; s3, performing topic modeling on the text message by using a biterm topic model to obtain a textvector based on the user message; s4, constructing a social network; s5, performing random walk on the social network to obtain a user identification sequence, and performing topic modeling by usinga biterm topic model to obtain a structural vector of the social network based on the user; and S6, splicing the text vector and the structure vector to serve as a feature vector of the current user,calculating similarity with other users, and taking k results with the highest similarity as user recommendation results. According to the method, the attributes of the user are analyzed, meanwhile, the similarity of invisible structures in the social network is considered, and through overall mastering of the whole social network, the method has the advantages of being high in accuracy, high in coverage rate, advanced in method and the like.

Description

technical field [0001] The invention relates to the technical field of social network in the Internet, in particular to a multi-dimensional user recommendation method and system for a social network. Background technique [0002] With the rapid development of the mobile Internet, especially the emergence of emerging social networking media such as Twitter, Facebook, and Weibo, a large number of users can make more friends while using these social networking platforms. Hobbies will resonate to varying degrees. Furthermore, since the Internet is not limited by space or time, it can grasp the latest information anytime and anywhere, and it has become an important channel for people to obtain information or share their experiences. Friend recommendation in social networks can recommend similar like-minded friends according to different user behaviors. [0003] Most of the existing user recommendation technologies are based on the user's own behavior, such as tweets, interactio...

AI Technical Summary

Problems solved by technology

[0003] Most of the existing user recommendation technologies are based on the user's own behavior, such as tweets, interaction with friends, etc. There is no user recommendation method that considers the implicit structure of social networks, so it is impossible to make more detailed recommendations

Moreover, the implicit expression in the network structure is very difficult for manual screening of features. Conventional methods cannot effectively mine the hidden network structure features, and cannot simultaneously consider the user's own attributes and the social network structure characteristics of the user.

Images