Graph contrast learning method oriented to recommendation system and based on hierarchical neighbor enhancement

By introducing a graph comparison learning method based on hierarchical neighbor enhancement in the recommendation system, the HNECL method can more accurately capture the complex relationship between the user and the item, solving the problem of insufficient supervision signals in the existing methods, and improving the accuracy and generalization ability of recommendations.

CN119646312AActive Publication Date: 2025-03-18TONGJI UNIV
5 Cites 1 Cited by

Patent Information

Application Number
CN202411873415.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-18
Publication Date
2025-03-18
Estimated Expiration
2044-12-18

AI Technical Summary

Technical Problem

The existing graph comparison learning methods fail to fully utilize the hierarchical structure and semantic neighbor relationships between users (or items) in the recommendation system, resulting in insufficient supervision signals and affecting the accuracy of recommendations.

Method used

A graph comparison learning method (HNECL) based on hierarchical neighbor enhancement is proposed. By constructing user-item interaction graph, user-user/item-item relationship graph, combined with graph neural network and hierarchical clustering algorithm, modeling the hierarchical structure and semantic neighbor representation of users (item) to construct joint optimization of comparison learning goals and recommended tasks.

Benefits of technology

Through more refined data augmentation strategies and contrast view construction, the HNECL method can more accurately capture the complex relationship between users and items, improve the accuracy and generalization capabilities of the recommendation system, and reduce the cost of data preparation.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention belongs to the field of recommendation systems, and particularly relates to a hierarchical neighbor enhancement-based graph contrast learning method oriented to a recommendation system. Comprising the following steps: constructing a user-article interaction graph, a user-user relation graph and an article-article relation graph according to user article interaction records; in combination with the graph neural network, node feature updating is carried out in a hierarchical fusion mode, and local and global structure neighbor representations of nodes are obtained; a hierarchical clustering algorithm is executed from bottom to top, and hierarchical semantic neighbor representation of nodes is modeled by using generated hierarchical prototype (centroid) features; respectively constructing hierarchical structure neighbors and hierarchical semantic neighbors of nodes according to the two types of neighbor representations, and optimizing user and article representations through back propagation; and generating a downstream recommendation result according to the optimized user and article characterization. According to the method, the accuracy of a recommendation system is improved, and the generalization ability of the model in a sparse data environment is enhanced.
Need to check novelty before this filing date? Find Prior Art