Online recommendation system based on deep learning and knowledge graph fusion

Abstract

The invention relates to an online recommendation system based on deep learning and knowledge graph fusion, and the system comprises a knowledge graph construction module which is configured to construct a knowledge graph according to a user entity, a product entity, a channel entity and the association relation among the entities; a model establishing module, which is configured to establish a deep learning model according to the knowledge graph, wherein the deep learning model comprises a deep learning unit and a linear sub-model; a recommendation module, which is configured to perform prediction by adopting the deep learning model according to the user entity, the product entity and / or the channel entity to obtain a recommendation prediction value, and complete online recommendation according to the recommendation prediction value; and a data transmission module and a client, which are configured to complete transmission and man-machine interaction of all data in the online recommendation system. According to the online recommendation system provided by the invention, products with diversity and relevance are provided for users, and the problems that recommended products are insufficient in freshness and too single due to data sparseness and cold start are solved.

Description

technical field [0001] The invention relates to the technical field of data mining and online recommendation, in particular to an online recommendation system based on fusion of deep learning and knowledge graph. Background technique [0002] With the rapid development of the Internet, all kinds of information are growing explosively, which leads to the emergence of "information overload". To solve this problem, the recommendation system came into being. The recommendation system can provide users with good decision support and personalized services. At present, collaborative filtering algorithms recommend products of interest to users according to their interests or preferences through methods such as data mining and machine learning. [0003] In the existing technology, most recommendation systems use expert rules or collaborative filtering algorithms to recommend products for users, but expert rules often cannot fully consider the relationship between users and recommend...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide an online recommendation system based on the fusion of deep learning and knowledge graphs to provide users with diverse and relevant products to solve the freshness of recommended products caused by data sparseness and cold start in the prior art Insufficient and too single problem

Images