Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RSEA-MVGNN: Multi-View Graph Neural Network with Reliable Structural Enhancement and Aggregation

Junyu Chen, Long Shi, Badong Chen

TL;DR

RSEA-MVGNN tackles the challenge of fusing multi-view graphs by coupling reliable structural enhancement with reliable aggregation. It uses subjective logic to estimate view-specific uncertainty and derives diversity through feature de-correlation, while aggregation parameters based on view beliefs and uncertainty let high-quality views dominate. The approach yields consistent gains over state-of-the-art baselines across five real-world datasets and demonstrates clear ablations showing the necessity of both components. The method offers a general, domain-agnostic framework for robust multi-view graph representation learning with practical efficiency advantages.

Abstract

Graph Neural Networks (GNNs) have exhibited remarkable efficacy in learning from multi-view graph data. In the framework of multi-view graph neural networks, a critical challenge lies in effectively combining diverse views, where each view has distinct graph structure features (GSFs). Existing approaches to this challenge primarily focus on two aspects: 1) prioritizing the most important GSFs, 2) utilizing GNNs for feature aggregation. However, prioritizing the most important GSFs can lead to limited feature diversity, and existing GNN-based aggregation strategies equally treat each view without considering view quality. To address these issues, we propose a novel Multi-View Graph Neural Network with Reliable Structural Enhancement and Aggregation (RSEA-MVGNN). Firstly, we estimate view-specific uncertainty employing subjective logic. Based on this uncertainty, we design reliable structural enhancement by feature de-correlation algorithm. This approach enables each enhancement to focus on different GSFs, thereby achieving diverse feature representation in the enhanced structure. Secondly, the model learns view-specific beliefs and uncertainty as opinions, which are utilized to evaluate view quality. Based on these opinions, the model enables high-quality views to dominate GNN aggregation, thereby facilitating representation learning. Experimental results conducted on five real-world datasets demonstrate that RSEA-MVGNN outperforms several state-of-the-art GNN-based methods.

RSEA-MVGNN: Multi-View Graph Neural Network with Reliable Structural Enhancement and Aggregation

TL;DR

RSEA-MVGNN tackles the challenge of fusing multi-view graphs by coupling reliable structural enhancement with reliable aggregation. It uses subjective logic to estimate view-specific uncertainty and derives diversity through feature de-correlation, while aggregation parameters based on view beliefs and uncertainty let high-quality views dominate. The approach yields consistent gains over state-of-the-art baselines across five real-world datasets and demonstrates clear ablations showing the necessity of both components. The method offers a general, domain-agnostic framework for robust multi-view graph representation learning with practical efficiency advantages.

Abstract

Graph Neural Networks (GNNs) have exhibited remarkable efficacy in learning from multi-view graph data. In the framework of multi-view graph neural networks, a critical challenge lies in effectively combining diverse views, where each view has distinct graph structure features (GSFs). Existing approaches to this challenge primarily focus on two aspects: 1) prioritizing the most important GSFs, 2) utilizing GNNs for feature aggregation. However, prioritizing the most important GSFs can lead to limited feature diversity, and existing GNN-based aggregation strategies equally treat each view without considering view quality. To address these issues, we propose a novel Multi-View Graph Neural Network with Reliable Structural Enhancement and Aggregation (RSEA-MVGNN). Firstly, we estimate view-specific uncertainty employing subjective logic. Based on this uncertainty, we design reliable structural enhancement by feature de-correlation algorithm. This approach enables each enhancement to focus on different GSFs, thereby achieving diverse feature representation in the enhanced structure. Secondly, the model learns view-specific beliefs and uncertainty as opinions, which are utilized to evaluate view quality. Based on these opinions, the model enables high-quality views to dominate GNN aggregation, thereby facilitating representation learning. Experimental results conducted on five real-world datasets demonstrate that RSEA-MVGNN outperforms several state-of-the-art GNN-based methods.
Paper Structure (23 sections, 15 equations, 4 figures, 2 tables, 1 algorithm)

This paper contains 23 sections, 15 equations, 4 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Multi-View Graph Neural Networks:
  4. Uncertainty-based Deep Learning:
  5. Our Method
  6. Preliminaries
  7. Multi-view Graph Representation Learning.
  8. Multi-view Graph Neural Network.
  9. Subjective Logic.
  10. Reliable Structural Enhancement and Aggregation
  11. Estimate Uncertainty of Intra-GNN.
  12. Reliable Structural Enhancement.
  13. Reliable Aggregation.
  14. Optimization
  15. Experiments
  16. ...and 8 more sections

Figures (4)

  • Figure 1: Visualization of the GSFs selected during structure enhancement. Squares of different colors represent nodes with different features in graphs. The red outlines indicate the enhanced nodes.
  • Figure 2: Illustration of RSEA-MVGNN. First, we learn view-specific beliefs and uncertainty as opinions. Based on the uncertainty, we apply reliable structural enhancement by feature de-correlation. Second, we construct aggregation parameters based on opinions of enhanced views, utilizing these parameters to facilitate high-quality views dominating inter-graph aggregation.
  • Figure 3: Ablation Study for RSEA-MVGNN.
  • Figure 4: Execution Time (seconds) and Space Requirements (gigabytes).