Recommendation method based on heterogeneous information network representation learning

Abstract

The invention discloses a recommendation method based on heterogeneous information network representation learning, wherein the recommendation method comprises the steps: extracting information, carrying out representation learning on nodes in a heterogeneous information network, wherein the nodes comprise user nodes and project nodes; obtaining low-dimensional vectors of a user and a project; directly docking the low-dimensional vectors of the user and the project with a recommendation task, inputting the low-dimensional vectors as recommended sample features into a domain perception factorization machine model, and carrying out feature selection by adding a group lasso as a regular item to complete score prediction between the user and the project; and completing recommendation according to score prediction. The heterogeneous information network representation learning method based on the meta-structure and the dynamic truncation random walk is adopted, not only can simple linear semantics be captured, but also mining of a complex nonlinear structure can be in order, and the problem of information loss caused by structural defects of a meta-path is effectively solved; and irreversible information loss possibly caused in an information fusion stage is avoided.

Description

technical field [0001] The invention relates to the technical field of recommendation, in particular to a recommendation method based on heterogeneous information network representation learning. Background technique [0002] In the era of big data, the recommendation system has become an indispensable tool for various online applications by virtue of its ability to provide users with instant and accurate personalized services. Collaborative filtering, which predicts user preferences based on similar users or items, is a popular and concerned recommendation algorithm in the field of recommender systems. Traditional collaborative filtering algorithms focus on mining user and item rating data, so there are inevitably many problems that affect recommendation performance. With the rapid development of information technology, additional data including user social relations, user or item metadata, user or item location, and item comments in the recommendation system become easy t...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a recommendation method based on heterogeneous information network representation learning, which is used to solve the problems in the prior art that the recommendation method has insufficient mining and does not consider the combined features, which affects the recommendation effect

Images