DGCR: depth graph convolution recommendation algorithm integrating social and residual

Haixin Huang; Tongyuan Wang

doi:10.1117/12.3025637

28 February 2024 DGCR: depth graph convolution recommendation algorithm integrating social and residual

Haixin Huang, Tongyuan Wang

Proceedings Volume 13071, International Conference on Mechatronic Engineering and Artificial Intelligence (MEAI 2023); 130712Q (2024) https://doi.org/10.1117/12.3025637
Event: International Conference on Mechatronic Engineering and Artificial Intelligence (MEAI 2023), 2023, Shenyang, China

Abstract

Graph neural network technology is widely used in social recommendation system to learn the embedded representation of users and items in the process of user interest graph and user social graph propagation. To solve the existing problems, this paper proposes a depth graph convolution recommendation algorithm (DGCR) that integrates social and residual errors. It uses lightweight graph convolution to carry out convolution propagation for users and items respectively in two graphs about users, and introduces the idea of residual network for deep propagation to prevent over-smoothing. At the same time, a more flexible multi-layer perceptron is used for image fusion to reduce the information gap. Compared with the baseline algorithm on LastFM and Ciao datasets, the experimental results show that DGCR algorithm has a significant improvement in recommendation effectiveness and cold start problems.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Haixin Huang and Tongyuan Wang "DGCR: depth graph convolution recommendation algorithm integrating social and residual", Proc. SPIE 13071, International Conference on Mechatronic Engineering and Artificial Intelligence (MEAI 2023), 130712Q (28 February 2024); https://doi.org/10.1117/12.3025637

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available