Bipartite graph-based live room recommendation method, related storage media and equipment

Abstract

The invention provides a live broadcast room recommendation method based on bipartite graphs. The method comprises the following steps of S1, establishing the bipartite graph between live broadcast rooms and users, and the bipartite graph between the live broadcast rooms and labels; S2, calculating a first contribution vector of the users for the live broadcast rooms according to the bipartite graph between the live broadcast rooms and the users, and a second contribution vector of the labels for the live broadcast rooms according to the bipartite graph between the live broadcast rooms and thelabels; S3, calculating a third contribution vector according to the first contribution vector and the second contribution vector; and S4, arranging elements according to a descending order accordingto values of the elements forming the third contribution vector, and recommending the live broadcast rooms corresponding to the elements which are arranged according to the descending order to targetusers. The invention also relates to a related computer readable memory medium and electronic equipment.

Description

technical field [0001] The invention relates to the field of big data recommendation, in particular to a bipartite graph-based live room recommendation method, related computer-readable storage media and electronic equipment. Background technique [0002] In the application field of big data, an important direction is to make personalized recommendations to users based on massive data. Currently the most popular recommendation methods are content-based, collaborative filtering-based, association-rule-based recommendation algorithms, and combinations of these recommendations. However, these algorithms also have some problems in practical applications, such as the sparse problem. When a user browses a system with few objects or rarely performs operations and evaluations, the user's behavior data will become very sparse. It is unreliable to use such data to calculate the similarity between users. In addition, according to the long-tail theory, some users may prefer some lesse...

AI Technical Summary

Problems solved by technology

However, these algorithms also have some problems in practical applications, such as the sparse problem. When a user browses a system with few objects or rarely performs operations and evaluations, the user's behavior data will become very sparse. It is unreliable to use such data to calculate the similarity between users

In addition, according to the long-tail theory, some users may prefer some lesser-known live broadcast rooms. Commonly used recommendation algorithms recommend more popular live broadcast rooms. will definitely like it, so it is impossible to accurately recommend the live broadcast room that the user really likes

Images