Project recommendation method of graph neural network based on directed and undirected structure information

Abstract

The invention discloses a project recommendation method of a graph neural network based on directed and undirected structure information, which comprises the following steps of: extracting undirected structure information in a graph by utilizing a graph convolutional network according to an adjacency relation of a session sequence graph, extracting directed structure information in the graph by utilizing a gated graph neural network, calculating to obtain an intermediate project implicit vector, and recommending the intermediate project implicit vector to the session sequence graph according to the intermediate project implicit vector. Performing linear transformation on the obtained intermediate item implicit vector to obtain a final item implicit vector; higher attention is distributed to repeatedly clicked items appearing in the session sequence, an attention mechanism is introduced when item implicit vectors are generated, and weight coefficients of corresponding items are modified according to the degree of dependency between the items. According to the method, the generated session vector is predicted more accurately in the recommendation process.

Description

technical field [0001] The invention relates to the technical field of recommendation methods, in particular to an item recommendation method based on a graph neural network with directed and undirected structural information. Background technique [0002] Recommender systems are one of the most important downstream applications in data mining and machine learning. It can help platform users alleviate the problem of information overload, and sort out valuable information among many web applications on e-commerce platforms, music websites, etc. In most recommender systems, the user's behavior sequence is arranged in time, and presents the characteristics of anonymity and large amount of data. In order to predict the user's behavior information at the next moment, the recommendation based on the session sequence learns the user's preferences by mining the sequential order feature information in the user's historical behavior. [0003] Session sequence refers to the sequence ...

AI Technical Summary

Problems solved by technology

[0008] The technical problem to be solved by the present invention is to provide an item recommendation method based on a graph neural network with directed and undirected structural information, so as to solve the problem of ignoring the repeated items in the click sequence existing in the existing technology and to solve the problem in generating items. The problem of not making better use of the structural information in the session sequence diagram when vector representation

Images