Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Dual-Optimized Adaptive Graph Reconstruction for Multi-View Graph Clustering

Zichen Wen, Tianyi Wu, Yazhou Ren, Yawen Ling, Chenhang Cui, Xiaorong Pu, Lifang He

TL;DR

A novel multi-view graph clustering method based on dual-optimized adaptive graph reconstruction, named DOAGC, which mainly aims to reconstruct the graph structure adapted to traditional GNNs to deal with heterophilous graph issues while maintaining the advantages of traditional GNNs.

Abstract

Multi-view clustering is an important machine learning task for multi-media data, encompassing various domains such as images, videos, and texts. Moreover, with the growing abundance of graph data, the significance of multi-view graph clustering (MVGC) has become evident. Most existing methods focus on graph neural networks (GNNs) to extract information from both graph structure and feature data to learn distinguishable node representations. However, traditional GNNs are designed with the assumption of homophilous graphs, making them unsuitable for widely prevalent heterophilous graphs. Several techniques have been introduced to enhance GNNs for heterophilous graphs. While these methods partially mitigate the heterophilous graph issue, they often neglect the advantages of traditional GNNs, such as their simplicity, interpretability, and efficiency. In this paper, we propose a novel multi-view graph clustering method based on dual-optimized adaptive graph reconstruction, named DOAGC. It mainly aims to reconstruct the graph structure adapted to traditional GNNs to deal with heterophilous graph issues while maintaining the advantages of traditional GNNs. Specifically, we first develop an adaptive graph reconstruction mechanism that accounts for node correlation and original structural information. To further optimize the reconstruction graph, we design a dual optimization strategy and demonstrate the feasibility of our optimization strategy through mutual information theory. Numerous experiments demonstrate that DOAGC effectively mitigates the heterophilous graph problem.

Dual-Optimized Adaptive Graph Reconstruction for Multi-View Graph Clustering

TL;DR

A novel multi-view graph clustering method based on dual-optimized adaptive graph reconstruction, named DOAGC, which mainly aims to reconstruct the graph structure adapted to traditional GNNs to deal with heterophilous graph issues while maintaining the advantages of traditional GNNs.

Abstract

Multi-view clustering is an important machine learning task for multi-media data, encompassing various domains such as images, videos, and texts. Moreover, with the growing abundance of graph data, the significance of multi-view graph clustering (MVGC) has become evident. Most existing methods focus on graph neural networks (GNNs) to extract information from both graph structure and feature data to learn distinguishable node representations. However, traditional GNNs are designed with the assumption of homophilous graphs, making them unsuitable for widely prevalent heterophilous graphs. Several techniques have been introduced to enhance GNNs for heterophilous graphs. While these methods partially mitigate the heterophilous graph issue, they often neglect the advantages of traditional GNNs, such as their simplicity, interpretability, and efficiency. In this paper, we propose a novel multi-view graph clustering method based on dual-optimized adaptive graph reconstruction, named DOAGC. It mainly aims to reconstruct the graph structure adapted to traditional GNNs to deal with heterophilous graph issues while maintaining the advantages of traditional GNNs. Specifically, we first develop an adaptive graph reconstruction mechanism that accounts for node correlation and original structural information. To further optimize the reconstruction graph, we design a dual optimization strategy and demonstrate the feasibility of our optimization strategy through mutual information theory. Numerous experiments demonstrate that DOAGC effectively mitigates the heterophilous graph problem.

Paper Structure

This paper contains 23 sections, 1 theorem, 24 equations, 5 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Multi-View Graph Clustering
  4. Heterophilous Graph Representation Learning
  5. Methodology
  6. Preliminaries
  7. Adaptive Graph Construction
  8. Dual Optimization Strategy for Reconstruction Graph
  9. View Weighting and Fusion
  10. Experiments
  11. Evaluation Setup and Metrics
  12. Datasets.
  13. Evaluation Metrics.
  14. Comparison Methods.
  15. Performance Comparison
  16. ...and 8 more sections

Key Result

lemma 1

DBLP:journals/corr/abs-2002-07017 Let $\mathbf{\tilde{X}_{\bar{y}_i}}$ and $\mathbf{\tilde{X}_{y_i}}$ be mutually redundant for $x_i$, i.e., the feature of node $i$ , where $\mathbf{\tilde{X}_{\bar{y}_i}}$ and $\mathbf{\tilde{X}_{y_i}}$ denote the nodes features with random masks belonging to differ where $\theta$ denotes the learnable parameters of the autoencoder. The proof is given in Appendix

Figures (5)

  • Figure 1: The framework of our DOAGC model. The inputs to each view are the node feature matrix $\mathbf{X}$ and the original adjacency matrix $\mathbf{A}$. The output is the consensus embedding H fused by each view node embedding h, after which H is used as input for k-means clustering.
  • Figure 1: Supplementary convergence analysis.
  • Figure 2: The adaptive process of $w$ on ACM.
  • Figure 3: Sensitive analysis of ACC and NMI on Wisconsin and Cornell with $order$ and $w$.
  • Figure 4: Convergence analysis.

Theorems & Definitions (1)

  • lemma 1