Social network user behavior prediction method based on quantified social influence

Abstract

The invention discloses a social network user behavior prediction method based on quantified social influence. The method comprises the following steps: data crawling, data preprocessing, network graph construction, user interesting extraction, inter-user influence quantification, modeling prediction, and model training evaluation. The prediction method is mainly used for predicting the user behavior such as tweeting behavior on the Twitter of the social network in the large-scale social network. Compared with the prior art, the method disclosed by the invention further considers the quantification of inter-user social influence in the social network, and the quantified influence is brought into a prediction model, so that the prediction model can sufficiently consider the influence on the user behavior by other users around a given user; and the user behavior is predicted from the angle of the user interests, the interests of the user are computed through the quantified social influence, and the user behavior is speculated through the user interests at last, and the method has higher accuracy.

Description

technical field [0001] The invention belongs to the technical field of data mining and behavior prediction, and more specifically relates to a method for predicting social network user behavior based on quantified social influence. Background technique [0002] With the development of Internet technology and mobile technology, social networks are emerging rapidly. At present, research on social networks mainly focuses on user behavior analysis and prediction, user relationship discovery, personalized content recommendation, community mining, topic detection and tracking, etc. The emergence of social networks such as Twitter has greatly accelerated the speed of information dissemination in the network. Users can quickly convey a blog post to more people by forwarding other people's blog posts and @other users, making the audience of a blog post geometrically Incrementally, resulting in the viral spread and diffusion of information. Since predicting user retweeting and @behav...

AI Technical Summary

Problems solved by technology

The characteristic factors can be expressed in various forms, such as user characteristics, social characteristics and Weibo characteristics. For a given user’s retweet prediction, most existing research focuses on user characteristics (user’s personal interests). In the selection of social features (influence among users), the impact of user interest on user retweet behavior will be mainly considered, that is, behavior prediction can be made directly by modeling user interest, or other users can be predicted by studying the influence among users. Whether a user will retweet a specific user's tweet, but in use, it is found that the prediction accuracy of the user's retweet behavior is low only based on user interest or influence among users

Images