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

Graph neural network recommendation method integrated with comment information

A neural network and recommendation method technology, applied in the field of computer information recommendation, can solve the problems of not being able to identify user concerns, single interactive information, and not being able to make good use of other useful information

Active Publication Date: 2021-04-13
ANHUI AGRICULTURAL UNIVERSITY
View PDF7 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] (1) Sensitive to data sparsity and cold start
[0006] (2) Single use of interaction information between users and items cannot make good use of other useful information
[0007] (3) For high-order neighbors, the user's concerns cannot be well identified

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
  • Graph neural network recommendation method integrated with comment information
  • Graph neural network recommendation method integrated with comment information
  • Graph neural network recommendation method integrated with comment information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0075]The invention is further illustrated below with reference to the accompanying drawings and examples.

[0076]Such asfigure 1 As shown in the present invention, the present invention is integrated into the comment information, using the Bert model from user comment text to extract the contents of the user and the item, then build two diagrams of the user and the item, using the diagram of the nerve network from the user - The structural expression of the learning user and items in the item two diagram, and fuses with the content expression as the final embedding representation of the user and the item. Finally, the multi-layer perceived machine is used to predict the user's interaction probability of the item, and sort the predictive interaction probability of each article, select a list of TOP-N recommended lists based on the user's prediction interaction probability of each item.

[0077]The invention includes the following steps:

[0078](1), such asfigure 2 As shown, the semantic fe...

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 graph neural network recommendation method integrated with the comment information. The method comprises steps of taking a user comment text as a source of node information, and carrying out the feature extraction of text data through a BERT model, so as to obtain a content expression vector of each node; a user-article bipartite graph being established according to interaction behaviors of a user and an article, structure expressions of the user and the article being learned by establishing first-order and third-order neighbor information of graph neural network aggregation nodes on the bipartite graph, and then the structure expressions and content expressions being fused to serve as final embedded expressions of the user and the article; and finally, predicting the interaction probability of the user for each article through a multilayer perceptron MLP. For the obtained prediction result, Top-N sorting is adopted to generate a recommended item list. According to the method, the user preference can be captured more accurately, the interest point of the user is found, and more accurate and effective recommendation is carried out on the interest point.

Description

Technical field[0001]The present invention relates to the field of computer information recommendation methods, and is specifically a method for integrating comment information.Background technique[0002]The recommended system can solve the information overload problem and develop quickly. For example, the software, shopping, music, video, etc. is related to recommendations, and the core is mostly based on the behavior of the user to study the hobbies of the user, and is interested in users from a large amount of data.[0003]Conventional recommendation algorithms have content-based recommendations based on the recommendation and mixing recommendations of collaborative filtration. The content of the content is used as information comparison filtering technology. The content of its items do not need to score, but it is necessary to extract the content characteristics and structural, and the recommended accuracy is not high and scalable. Collaborative filtration adopts the approval of in...

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
IPC IPC(8): G06F16/9535G06F40/284G06F40/30G06N3/04G06N3/08
CPCG06F16/9535G06N3/084G06F40/284G06F40/30G06N3/047G06N3/045
Inventor 吴国栋王伟娜李景霞涂立静刘玉良
Owner ANHUI AGRICULTURAL UNIVERSITY
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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com