Method, device and storage medium for recommending media data

Abstract

The invention discloses a media data recommendation method, device and storage medium, belonging to the technical field of the Internet. The method includes: obtaining the user characteristics of the target user; according to the multiple media data to be recommended, obtaining a plurality of candidate media data matching the characteristics of the interest tag; for each candidate media data in the multiple candidate media data, the Combining the user characteristics with the data characteristics of the candidate media data to obtain the combined associated features, and obtaining the degree of association of the associated features based on the first recommendation model; according to the degree of association corresponding to multiple media candidates, determine the recommended Target media data. By screening out multiple candidate media data that match the characteristics of the interest tags, the amount of calculation can be reduced, and the recommendation efficiency can be improved. According to the association characteristics and the first recommendation model, the degree of association can be determined, and the personalized recommendation can be effectively targeted at the target user, improving the efficiency of the recommendation. The accuracy of the correlation degree improves the recommendation effect.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to a media data recommendation method, device and storage medium. Background technique [0002] With the rapid development and widespread popularization of Internet technology, a variety of media data can be disseminated on the Internet, providing users with a variety of rich information. If the disseminated media data is directly displayed to the user, the amount of data will be too much, and it will be difficult for the user to find the media data they are interested in. Therefore, according to the user's preference for different media data, we can recommend the media data that the user is interested in. media data. [0003] In related technologies, multiple categories of media data and the weight of each category are preset, and the multiple categories have a hierarchical relationship, which means that a category at a previous level includes multiple subcategories at a lowe...

AI Technical Summary

Problems solved by technology

[0004] The above scheme needs to obtain the correlation degree of each media data in multiple media data, and the calculation amount is relatively large

Moreover, the weight of each category is pre-set, and the setting method is not flexible enough to effectively make personalized recommendations for each user, resulting in inaccurate correlations and poor recommendation effects

Images