Supercharge Your Innovation With Domain-Expert AI Agents!

Cross-domain sequence recommendation method and system based on domain perception graph convolutional neural network

A convolutional neural network and recommendation method technology, applied in the field of cross-domain sequence recommendation methods and systems, can solve problems such as ignoring explicit structural information, limiting the ability to learn user and item representation, and limited ability to complex relationships

Pending Publication Date: 2021-12-03
SHANDONG NORMAL UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the vast majority of these RNN-based methods aim at discovering sequential dependencies and are limited in their ability to capture complex relationships between related entities (i.e., users and items in two domains).
Therefore, this limits the ability to learn user and item representations and ignores explicit structural information connecting the two domains, e.g., item-user-item paths

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
  • Cross-domain sequence recommendation method and system based on domain perception graph convolutional neural network
  • Cross-domain sequence recommendation method and system based on domain perception graph convolutional neural network
  • Cross-domain sequence recommendation method and system based on domain perception graph convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0038] Such as figure 1 As shown, Embodiment 1 of the present disclosure provides a cross-domain sequence recommendation method based on a domain-aware graph convolutional neural network, a cross-domain sequence recommendation method based on a domain-aware graph convolutional neural network, including:

[0039] Obtain the behavior sequence of accounts under different domains;

[0040] Build a domain-aware graph convolutional neural network DA-GCN;

[0041] Based on DA-GCN, item recommendation is made according to the user's historical behavior;

[0042] Wherein, the construction of a domain-aware graph convolutional neural network includes: constructing a cross-domain sequence (CDS) graph, converting the mixed sequence of shared accounts into a CDS graph; node representation learning, updating all users and users through information transfer and information aggregation Representation of items; feature combination of sequence embedding and user embedding splicing; use predic...

Embodiment 2

[0121] Embodiment 2 of the present disclosure provides a cross-domain sequence recommendation system based on a domain-aware graph convolutional neural network, including:

[0122] a sequence conversion module configured to convert a mixed sequence of shared accounts into a CDS graph;

[0123] A node representation module, configured to learn node embeddings for users and items by passing information from connected neighboring nodes.

[0124] The feature combining module is configured to concatenate sequence embeddings with user embeddings and input to the prediction layer.

[0125] The prediction layer module is configured to train DA-GCN with a negative log-likelihood loss function.

[0126] The working method of the system is the same as the cross-domain sequence recommendation method based on the domain-aware graph convolutional neural network provided in Embodiment 1, and will not be repeated here.

Embodiment 3

[0128] Embodiment 3 of the present disclosure provides a medium on which a program is stored, and when the program is executed by a processor, the method of cross-domain sequence recommendation based on a domain-aware graph convolutional neural network as described in Embodiment 1 of the present disclosure is implemented. step,

[0129] The detailed steps are the same as the cross-domain sequence recommendation method based on the domain-aware graph convolutional neural network provided in Embodiment 1, and will not be repeated here.

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 provides a cross-domain sequence recommendation method based on a domain perception graph convolutional neural network. The method comprises the following steps: constructing a cross-domain sequence (CDS) graph, specifically, converting a mixed sequence of a shared account into the CDS graph; representation learning of the nodes, including representation learning of the user and representation learning of a user-specific item; and carrying out feature combination on sequence embedding and user embedding splicing to obtain a prediction layer. The information transmitted by the connected nodes is aggregated by using the message transmission strategy in the domain awareness graph convolutional network, the node embedding of the user and the project is learned by transmitting the information from the connected adjacent nodes, two attention mechanisms are provided, and the related information is discriminatively selected in the message transmission process. Therefore, multi-aspect interaction can be modeled, and fine-grained domain knowledge is transmitted by considering structural information.

Description

technical field [0001] The present disclosure relates to the field of convolutional neural networks, in particular to a cross-domain sequence recommendation method and system based on domain-aware graph convolutional neural networks. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art. [0003] Cross-Domain Sequential Recommendation (CSR) is a recommendation task that aims to leverage users' historical interactions from multiple domains to recommend the next item. Nowadays, CSR has received extensive research attention because users need to register with different platforms to obtain domain-specific services, such as music subscription and food delivery. In this work, an emerging but challenging scenario, Shared Account Cross-Domain Sequential Recommendation (SCSR), is studied, where multiple users share an account and their interactions are recorded...

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): G06F16/9535G06N3/04G06N3/08
CPCG06F16/9535G06N3/084G06N3/045
Inventor 郭磊唐丽王新华
Owner SHANDONG NORMAL UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com