A method, device and storage medium for information recommendation

Abstract

The present application provides an information recommendation method, device and storage medium, which relate to cloud technology and the field of computer technology, and are used to improve scoring accuracy. In this application, after receiving the information scoring request, send a request message carrying the identification of the scoring service model to the online parameter server, so as to pull the model parameters and model network structure of the scoring service model from the online parameter server, wherein the online parameter server The model parameters and model network structure of the trained scoring service model are stored in the middle distribution; based on the model parameters and model network structure, the trained scoring service model is determined; based on the trained scoring service model, the information scoring request for recommendation The information is scored, and the obtained scoring result is returned to the upstream server, so that the upstream server recommends the information to be recommended to the target user according to the scoring result.

Description

Technical field [0001] The present application relates to the field of computer technology, there is provided an information recommendation method, apparatus and a storage medium. Background technique [0002] For information recommendation, is carried out by a machine learning model scoring information, sort information based on scoring results, the recommended information to the user in accordance with the order, so that users get the information of interest. Therefore Recommended process requires training machine learning models used to sort various links and recommended recall aspects of information systems. Among them, sorting and recall of these two aspects is the core of the information to rate the service that is the most important scoring information for the recommended model. [0003] Currently, all the parameters scoring service model needs to be loaded into a stand-alone service that the content loaded into scoring server, but when the memory of all parameters of the ...

AI Technical Summary

Problems solved by technology

[0003] At present, all parameters of the scoring service model need to be loaded into the stand-alone service, that is, loaded into the content scoring server, but when all the parameters of the model reach the upper memory limit of the content scoring server, the parameters of the scoring service model cannot be loaded to the content scoring server , causing the content scoring server to fail to score the information;

[0004] And the model parameters and model network structure of the new scoring service model generated by each training will change, and the new scoring service model generated by training will be uploaded to the online service in the form of a model file, and the online service will download the model file and obtain The model parameters and model network structure in the model file are merged to generate a model. It takes time to store and download the model file in the network, so there is a certain delay in the update of the model, resulting in a decrease in scoring accuracy

Images