Efficient Graph Optimization via Distance-Aware Graph Representation Learning

2024/06/25 by Dong Liu, Liu, Dong, Yu, Yanxuan · 2 citations
Computer Science · Mathematics · #Artificial intelligence #Artificial neural network #Combinatorics #Computer science #Graph #Mathematics #Neural Networks and Applications #Theoretical computer science #Topology (electrical circuits)

paper · pdf · doi:10.48550/arxiv.2406.17281

published in arXiv (Cornell University) (Cornell University)

openalex publication_date 2024/06/25 · openalex created_date 2024/06/27 · openalex updated_date 2026/08/08

Abstract

We propose an efficient framework that integrates distance-aware multi-hop message passing with dynamic topology refinement. Unlike standard GNNs that rely on shallow, fixed-hop aggregation, DRTR leverages both static preprocessing and dynamic resampling to capture deeper structural dependencies. A Distance Recomputator prunes semantically weak edges using adaptive attention, while a Topology Reconstructor establishes latent connections among distant but relevant nodes. This joint mechanism enables more expressive and robust graph representation optimization across evolving graph structures. Extensive experiments demonstrate that DRTR outperforms baseline GNNs in both accuracy and scalability, with at most 20% computational overhead, especially in complex and noisy graph environments.

Cited by

• Anomaly Detection and Early Warning Mechanism for Intelligent Monitoring Systems in Multi-Cloud Environments Based on LLM

Related

• Distance Encoding: Design Provably More Powerful Neural Networks for Graph Representation Learning
• Distance and Hop-wise Structures Encoding Enhanced Graph Attention Networks
• GEN: A Practical Alternative to Graph Transformers for Long-Range Graph Modeling
• Template based Graph Neural Network with Optimal Transport Distances
• Recurrent Distance Filtering for Graph Representation Learning