Information recommendation method based on graph convolution and neural collaborative filtering

Abstract

The invention discloses an information recommendation method based on graph convolution and neural collaborative filtering. In combination with the advantages of a graph convolution neural network model, fusion processing can be carried out on various information in an intuitive manner, so that not only feature information of a user but also attribute information of the user can be received, and relatively good recommendation performance can be achieved for sparse score data; and input and parameters of the model are subjected to optimization modeling by using multiple skills, so that the detail problems encountered possibly are solved. In addition, a nonlinear neural network-based collaborative filtering method is introduced as a decoder part of the model, so that user and article codes output by a graph convolution encoder can be well utilized, and through an end-to-end model, all processes run in the same framework without being trained separately. Through the processing of input data and the training and prediction of the model, a complete score prediction matrix can be obtained.

Description

Technical field [0001] The present invention relates to the technical field of information recommendation, in particular to an information recommendation method based on graph convolution and neural collaborative filtering. Background technique [0002] Because many online services on the Internet can give users a wealth of choices, the provision of highly accurate and personalized recommendation results has been generally regarded as the cornerstone of many important Internet applications, such as the recommendation of machine learning methods in entertainment, shopping, and academic fields. System (RS, RecommenderSystem). For individuals, the use of recommendation systems allows users to obtain and use information in a more effective way. In addition, many companies have widely used the technology of recommendation systems to locate target users by actively recommending products or services. With the continuous progress of various machine learning and data mining technologies...

AI Technical Summary

Problems solved by technology

In recent years, many deep learning-based methods try to fuse auxiliary information for recommendation, and have achieved performance improvements compared to traditional methods, but few models can handle graph structure information

The existing collaborative filtering work based on graph convolution is still based on the traditional matrix decomposition method, using the method of linear inner product calculation to combine the encoding vectors, and predicting ratings in this way will also lead to a decrease in recommendation performance

However, the existing nonlinear neural collaborative filtering (NCF, NEURAL COLLABORATIVE FILTERING) uses a nonlinear neural network for the collaborative filtering process, but the auxiliary information is not considered in the model recommendation process, so it cannot make good use of auxiliary information and scoring. Information makes predictions on unknown ratings

The above problems have restricted the development of recommender system models.

Images