GravityGraphSAGE: Link Prediction in Directed Attributed Graphs

2026-05-10

AI Summary

This work addresses the limitations of existing link prediction methods, which are predominantly confined to undirected graphs and often neglect node attributes. To overcome these shortcomings, the authors propose an efficient link prediction framework tailored for directed attributed graphs. Their approach innovatively adapts the GraphSAGE backbone to directed graph settings and introduces a gravity-inspired decoder that explicitly models the joint influence of edge directionality and node attributes. Extensive experiments on Cora, Citeseer, PubMed, and 16 real-world graph datasets demonstrate that the proposed method significantly outperforms current state-of-the-art graph deep learning models for link prediction, achieving both superior predictive performance and strong scalability.

Abstract

Link prediction (inferring missing or future connections between nodes in a graph) is a fundamental problem in network science with widespread applications in, e.g., biological systems, recommender systems, finance and cybersecurity. The ability to accurately predict links has significant real-world applications, such as detecting fraudulent financial transactions or identifying drug-target interactions in biomedicine. Despite a rich literature, link prediction is still challenging, especially for graphs enriched with information on edges (direction) and nodes (attributes). In fact, research on link prediction, especially the one based on Graph Deep Learning (GDL), has mostly focused on undirected graphs, without fully leveraging node attributes. Here, we fill this gap by proposing Gravity-GraphSAGE (GG-SAGE), a modified version of GraphSAGE, a GDL model for node embeddings, composed of a gravity-inspired decoder. This implementation is the first example in the literature of a GraphSAGE backbone adopted for directed link prediction. Using the benchmark datasets Cora, Citeseer, PubMed and 16 real-world graphs from the online Netzschleuder repository, we show that our proposed model outperforms state-of-the-art GDL link prediction techniques. Using further experimental evidence, we relate the quality of the output of our model with various characteristics of the graph, suggesting that our framework scales well when applied to data of increasing complexity.

Problem

• link prediction
• directed attributed graphs
• Graph Deep Learning
• node attributes
• edge direction

Innovation

• directed attributed graphs
• link prediction
• GraphSAGE
• gravity-inspired decoder
• graph deep learning