Random Walk Diffusion for Efficient Large-Scale Graph Generation

2024-08-08

arXiv.org

AI Summary

Existing diffusion-based graph generation methods suffer from poor scalability to large-scale graphs due to high computational overhead and superlinear time complexity. To address this, we propose ARROW-Diff—a stochastic-walk-driven autoregressive diffusion framework that avoids global modeling via iterative local sampling and dynamic graph pruning, achieving linear-time complexity O(|E|). Its core innovation lies in decoupling the diffusion process into edge-level autoregressive generation and structure-aware pruning, eliminating reliance on complex GNN architectures. Evaluated on multiple large-scale benchmark graphs, ARROW-Diff achieves an average 3.2× speedup in generation time over prior methods while simultaneously improving key topological metrics—including degree distribution, clustering coefficient, and average path length—outperforming all state-of-the-art baselines across the board.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Computer Vision: Diffusion Models for Vision
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Web Mining and Content Analysis: Content-based information diffusion

Abstract

Graph generation addresses the problem of generating new graphs that have a data distribution similar to real-world graphs. While previous diffusion-based graph generation methods have shown promising results, they often struggle to scale to large graphs. In this work, we propose ARROW-Diff (AutoRegressive RandOm Walk Diffusion), a novel random walk-based diffusion approach for efficient large-scale graph generation. Our method encompasses two components in an iterative process of random walk sampling and graph pruning. We demonstrate that ARROW-Diff can scale to large graphs efficiently, surpassing other baseline methods in terms of both generation time and multiple graph statistics, reflecting the high quality of the generated graphs.

Problem

• Generating graphs resembling real-world data distribution
• Scaling diffusion methods for large graph generation
• Improving efficiency and quality in graph generation

Innovation

• Random walk-based diffusion for graph generation
• Iterative random walk sampling and graph pruning
• Efficient scaling to large graphs