Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Democratic Recommendation with User and Item Representatives Produced by Graph Condensation

Jiahao Liang, Haoran Yang, Xiangyu Zhao, Zhiwen Yu, Guandong Xu, Wanyu Wang, Kaixiang Yang

TL;DR

DemoRec is a framework that leverages graph condensation to generate user and item representatives for recommendation tasks that significantly reduces graph size and computational complexity and mitigates the over-reliance on high-order information, a critical challenge in large-scale bipartite graphs.

Abstract

The challenges associated with large-scale user-item interaction graphs have attracted increasing attention in graph-based recommendation systems, primarily due to computational inefficiencies and inadequate information propagation. Existing methods provide partial solutions but suffer from notable limitations: model-centric approaches, such as sampling and aggregation, often struggle with generalization, while data-centric techniques, including graph sparsification and coarsening, lead to information loss and ineffective handling of bipartite graph structures. Recent advances in graph condensation offer a promising direction by reducing graph size while preserving essential information, presenting a novel approach to mitigating these challenges. Inspired by the principles of democracy, we propose \textbf{DemoRec}, a framework that leverages graph condensation to generate user and item representatives for recommendation tasks. By constructing a compact interaction graph and clustering nodes with shared characteristics from the original graph, DemoRec significantly reduces graph size and computational complexity. Furthermore, it mitigates the over-reliance on high-order information, a critical challenge in large-scale bipartite graphs. Extensive experiments conducted on four public datasets demonstrate the effectiveness of DemoRec, showcasing substantial improvements in recommendation performance, computational efficiency, and robustness compared to SOTA methods.

Democratic Recommendation with User and Item Representatives Produced by Graph Condensation

TL;DR

DemoRec is a framework that leverages graph condensation to generate user and item representatives for recommendation tasks that significantly reduces graph size and computational complexity and mitigates the over-reliance on high-order information, a critical challenge in large-scale bipartite graphs.

Abstract

The challenges associated with large-scale user-item interaction graphs have attracted increasing attention in graph-based recommendation systems, primarily due to computational inefficiencies and inadequate information propagation. Existing methods provide partial solutions but suffer from notable limitations: model-centric approaches, such as sampling and aggregation, often struggle with generalization, while data-centric techniques, including graph sparsification and coarsening, lead to information loss and ineffective handling of bipartite graph structures. Recent advances in graph condensation offer a promising direction by reducing graph size while preserving essential information, presenting a novel approach to mitigating these challenges. Inspired by the principles of democracy, we propose \textbf{DemoRec}, a framework that leverages graph condensation to generate user and item representatives for recommendation tasks. By constructing a compact interaction graph and clustering nodes with shared characteristics from the original graph, DemoRec significantly reduces graph size and computational complexity. Furthermore, it mitigates the over-reliance on high-order information, a critical challenge in large-scale bipartite graphs. Extensive experiments conducted on four public datasets demonstrate the effectiveness of DemoRec, showcasing substantial improvements in recommendation performance, computational efficiency, and robustness compared to SOTA methods.

Paper Structure

This paper contains 30 sections, 27 equations, 6 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Problem Statement
  3. Graph Condensation.
  4. Graph Recommendation.
  5. Framework
  6. Overall Architecture
  7. Graph Condensation Stage
  8. Initialization
  9. Relay Model Training
  10. Alignment Optimization
  11. Recommendation Stage
  12. Model Optimization
  13. Time and Space Complexity Analysis
  14. Space Complexity
  15. Time Complexity
  16. ...and 15 more sections

Figures (6)

  • Figure 1: A toy example of the democratic election process.
  • Figure 2: The detailed architecture of the proposed DemoRec. The architecture consists of two distinct parts: the Graph Condensation module and the Recommendation module.
  • Figure 3: Performance comparison between DemoRec and Non-Bi DemoRec on Precision, Recall, and NDCG.
  • Figure 5: Component Analysis with three backbone GNNs.
  • Figure 6: Hyper-parameter study of hidden space dimensions and condensation ratio $\alpha$.
  • ...and 1 more figures