Content recommendation method and device

Abstract

The embodiment of the invention provides a content recommendation method and device. The method comprises the steps that recent behaviors of users to be recommended, long-term interests of the users to be recommended in portraits of the users to be recommended and user characteristics of the users to be recommended are combined to serve as input of a trained interest representation model, and vector embedding representation of overall interests of the users to be recommended is output through the trained interest representation model. Because the long-term interest is the interest in the historical time period and the recent behavior is the recent behavior before the current time, the long-term interest is more stable than the recent behavior, and the recent behavior becomes a part of thelong-term interest along with the advancement of time. Therefore, the recent behaviors of the to-be-recommended users, the long-term interests of the to-be-recommended users in the to-be-recommended user portraits and the user characteristics of the to-be-recommended users are considered to be combined, the interests of the users can be described more accurately and comprehensively, and then interested contents are recommended to the to-be-recommended users based on the vector embedding representation of the whole to-be-recommended users.

Description

technical field [0001] The present invention relates to the technical field of personalized recommendation services, in particular to a content recommendation method and device. Background technique [0002] With the development of network technology, the amount of information is increasing, and users are often helpless in the face of this information. Therefore, it is necessary to distinguish these information through the network to provide content services for users. In this context, Personalized Recommender Services (PRS for short) technology emerges as the times require. Personalized recommendation in PRS is an important way of content distribution. By describing user interests, it recommends interesting content for users, so as to realize content distribution. [0003] The main steps of the content recommendation method in the related art are as follows: [0004] The first step is to obtain the vector embedding representation of all items as a distributed representati...

AI Technical Summary

Problems solved by technology

[0010] The purpose of the embodiments of the present invention is to provide a method and device for content recommendation, which is used to solve the problem that the prior art cannot accurately reflect the user's interest in the preset time period before the current time due to the relatively large fluctuations in the recent behavior changes, and further provide users with Recommend content of interest Not accurate enough technical issues

Images