Hot recall method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a popularity recall method and device, electronic equipment and a storage medium. The method comprises the steps: setting a plurality of popularity levels forall to-be-recommended information to be recalled; according to the popularity score of each part of to-be-recommended information, dividing each part of to-be-recommended information into the corresponding popularity level; determining a recall probability corresponding to each popularity level, and recalling the to-be-recommended information included in the popularity level according to the corresponding recall probability to obtain a recall result; and outputting the recall result. Therefore, according to the embodiment of the invention, the to-be-recommended information with a relatively low popularity score also has a certain exposure probability; Meanwhile, some randomness is introduced, and to-be-recommended information with the same popularity cannot be recommended to different users, so that multi-level popularity recall in the recommendation system is realized, recommendation diversification is enriched, and individual requirements of different recommendation contents are met.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a heat recall method, device, electronic equipment and storage medium. Background technique [0002] As the amount of information increases, various information services are using various recommendation methods to present more valuable content to users. Among them, popularity-based recommendation, as one of the recommendation methods, is widely used in multiple scenarios such as hot information push, boutique list, cold start recommendation, and interest exploration. [0003] At present, the popularity-based recommendation process can generally be roughly divided into two main stages: popularity score calculation and popularity recall. Among them, in the popularity recall stage, a list is generally generated in reverse order according to the popularity value of the information to be recommended. When recalling, the information to be recommended is obtained in order of po...

AI Technical Summary

Problems solved by technology

[0004] However, in the prior art, it can only be recalled in order of popularity, and the less popular information to be recommended cannot be exposed before the highly popular information to be recommended is exposed; at the same time, due to the sequential recall, different users see the The order of the recommended information is the same, resulting in problems such as the single way of popular recall and the same recommended content

Images