Hybrid recommendation system and method based on multi-objective optimization

Abstract

The invention discloses a mixed recommendation system and method based on multi-objective optimization. The method comprises the following steps: acquiring data required by a recommendation system and establishing a corresponding data set; generating a mixed recommendation algorithm candidate set according to a plurality of indexes such as accuracy, novelty, diversity and recall rate of recommendation by using a mixed recommendation method; modeling the plurality of evaluation indexes into a multi-objective optimization problem, and converting a recommendation problem of the plurality of evaluation indexes into a high-dimensional multi-objective optimization problem; and improving a multi-objective optimization method, and selecting a recommendation set which most meets user requirements from the plurality of mixed recommendation algorithm candidate sets. According to the invention, the multi-objective optimization mixed recommendation method is applied to the recommendation system, so that a plurality of recommendation results can be recommended to the user more accurately, and the demand of the recommendation results of the user is better met.

Description

technical field [0001] The invention relates to the field of multi-objective optimization, in particular to a multi-objective optimization-based hybrid recommendation system and a method thereof. Background technique [0002] With the rapid development of Internet technology, the scale of network information has grown exponentially. The vast and complex network information allows us to obtain more diverse information, but how to allow users to accurately obtain the information they want from the massive information has become a difficult problem. From the perspective of information providers, how to accurately availability to users who are interested in them is also a challenge. In this case, recommender systems come into being. [0003] The recommender system does not require users to express their needs clearly, but presents some information that users may be interested in by analyzing users' historical behaviors. In recent years, recommendation systems have shown good ...

AI Technical Summary

Problems solved by technology

A single different recommendation algorithm has shown good performance in different applications. Traditional recommendation algorithms only use the accuracy of the recommendation as the only measure of the recommendation results when making recommendations to users, although this recommendation method can accurately match The user's interest needs, but the type of recommendation is relatively single, and it is easy for users to fall into a single interest circle

Images