Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Collaborative filtering recommendation method based on enhanced graph learning

A collaborative filtering recommendation and product technology, applied in complex mathematical operations, instruments, data processing applications, etc., can solve the problems of lack of feature information of user nodes and product nodes, difficulty in learning graphs, etc., to improve recommendation accuracy and scalability. , the effect of expanding the data dimension

Active Publication Date: 2021-06-04
HEFEI UNIV OF TECH
View PDF5 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the recommendation system based on collaborative filtering, user nodes and product nodes lack feature information, and the traditional graph reconstruction method based on feature points is difficult to work, which brings difficulties to graph learning

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Collaborative filtering recommendation method based on enhanced graph learning
  • Collaborative filtering recommendation method based on enhanced graph learning
  • Collaborative filtering recommendation method based on enhanced graph learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0088] In order to verify the effectiveness of this method, the present invention uses three public data sets commonly used in recommendation systems: Movielens-1M, Amazon-Video Games and Pinterest to conduct experiments. For each dataset, users with less than five scoring records were screened to obtain the final dataset used.

[0089] The present invention adopts Hit Ratio (HR) and Normalized Discounted Cumulative Gain (NDCG) as evaluation criteria. The present invention uses 7 methods to compare the effects, namely: BPR, NGCF, LR-GCCF, LightGCN, GAT, DropEdge, and GLCN.

[0090] Table 1 The recommendation effect of the method of the present invention and the comparison method on the Movielens-1M data set

[0091] Models HR@5 HR@10 HR@15 HR@20 NDCG@5 NDCG@10 NDCG@15 NDCG@20 BPR 0.1495 0.2006 0.2454 0.2894 0.1363 0.1552 0.1713 0.1857 NGCF 0.1548 0.2106 0.2602 0.3011 0.1415 0.1621 0.1795 0.1929 LR-GCCF 0.1593 0.2116 ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a collaborative filtering recommendation method based on enhanced graph learning, and the method comprises the steps: 1, constructing a bipartite graph of a user to a product, wherein the bipartite graph comprises a user node set, a product node set and an adjacent matrix; 2, obtaining an embedded matrix as a node feature through one-hot coding; 3, calculating a similar matrix according to the current node characteristics, and performing sparsification; 4, taking the sparse similar matrix as a residual item and adding the residual item to the adjacent matrix to obtain an enhanced adjacent matrix; 5, carrying out feature propagation according to the enhanced adjacent matrix structure graph convolution layer, and obtaining node representation; and 6, obtaining a score matrix from the prediction layer according to the node representation, thereby realizing product recommendation. According to the method, the graph structure information can be adaptively learned based on the similarity between the nodes, and the robustness and integrity of the graph are enhanced, so that more accurate node representation is learned, and the recommendation performance is improved.

Description

technical field [0001] The invention relates to the field of personalized recommendation, in particular to a collaborative filtering recommendation method based on enhanced graph learning. Background technique [0002] In the Internet era of information explosion, information overload has become a problem that restricts users from effectively obtaining external information. The recommendation system aims to mine the historical behavior of users and recommend products that meet their interests and preferences to help users obtain the desired information from massive data. The recommendation model based on collaborative filtering is one of the most mainstream recommendation systems, and its modeling of users' potential interests can make personalized recommendations. Collaborative filtering models are widely used in recommendation scenarios, but their performance is limited by the sparsity of data. [0003] The graph-based collaborative filtering model models the user-produc...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/9535G06F16/9536G06Q50/00G06F17/16
CPCG06F16/9535G06F16/9536G06Q50/01G06F17/16
Inventor 吴乐杨永晖张琨汪萌洪日昌
Owner HEFEI UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products