Method and system for mining interests of social network users

Abstract

The invention discloses a social network user interest mining method. The method is characterized by comprising the steps: rearranging user groups which are divided aiming at respective concerned objects of all users in a social network, so as to obtain all user groups in the social network; selecting a user group for each preset interest category, and enabling the selected user group to serve as an interest core group corresponding to the related interest category; calculating the similarity between the other user groups and each selected interest core group, and enabling the user groups with similarity higher than a set threshold value to serve as a generalized interest group of the related interest core group; determining an interest mining result for each user in the social network according to the interest categories corresponding to the generalized interest group containing the related user. By applying the method, the accuracy of interest mining can be improved.

Description

technical field [0001] The present invention relates to Internet technology, in particular to a method and device for mining interests of social network users. Background technique [0002] With the development of Internet technology, there is a sharp contrast between the massive amount of information on the Internet and the ability of users to effectively obtain information. In order to improve users' satisfaction in finding information, in social network SNS (Social Network Service, social network service), personalized information services that provide targeted information according to users' interests came into being, and the basis of personalized information services It is the mining of user interests. [0003] In practical applications, the behavior of users on social networks is generally carried out actively, for example, defining or selecting tags by themselves, browsing pages, using on-site products or third-party apps, publishing blog posts or liking or favoritei...

AI Technical Summary

Problems solved by technology

However, custom tags are generated or selected by users. In practical applications, some tags are too generalized to be classified into interest categories. There are also cases where users do not tag or tag at will.

Therefore, if the interest mining method based on user tags is used, the accuracy of mining user interests is not high only based on user-defined tags.

Moreover, for browsing users with less active behavior, there may be no custom tags, so the interest mining method based on user tags cannot mine the interests of these browsing users.

[0006] To sum up, the existing interest mining methods of social network users are not very accurate, especially for the interests of browsing users with less active behavior.

Images