Edge Prompt Tuning for Graph Neural Networks

2025-03-02

AI Summary

Existing graph neural network (GNN) pretraining suffers from objective misalignment with downstream tasks, while mainstream graph prompt-tuning methods operate solely on nodes, neglecting edge structural information and thereby limiting representation quality. To address this, we propose **EdgePrompt**, the first edge-level prompt-tuning framework, which injects learnable edge prompt vectors into the message-passing process to explicitly model edge dependencies. EdgePrompt is model- and pretraining-agnostic, ensuring broad applicability across diverse GNN architectures and pretraining paradigms. We further provide theoretical convergence analysis for both node- and graph-classification tasks. Extensive experiments across 10 benchmark graph datasets and 4 pretraining strategies demonstrate that EdgePrompt consistently outperforms six state-of-the-art baselines. The implementation is publicly available.

Technology Category

Machine Learning: Graph-based Machine LearningNatural Language Processing: Prompt Engineering / PromptingData 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 graphsSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsWeb Mining and Content Analysis: Large pretrained models with web data

Abstract

Pre-training powerful Graph Neural Networks (GNNs) with unlabeled graph data in a self-supervised manner has emerged as a prominent technique in recent years. However, inevitable objective gaps often exist between pre-training and downstream tasks. To bridge this gap, graph prompt tuning techniques design and learn graph prompts by manipulating input graphs or reframing downstream tasks as pre-training tasks without fine-tuning the pre-trained GNN models. While recent graph prompt tuning methods have proven effective in adapting pre-trained GNN models for downstream tasks, they overlook the crucial role of edges in graph prompt design, which can significantly affect the quality of graph representations for downstream tasks. In this study, we propose EdgePrompt, a simple yet effective graph prompt tuning method from the perspective of edges. Unlike previous studies that design prompt vectors on node features, EdgePrompt manipulates input graphs by learning additional prompt vectors for edges and incorporates the edge prompts through message passing in the pre-trained GNN models to better embed graph structural information for downstream tasks. Our method is compatible with prevalent GNN architectures pre-trained under various pre-training strategies and is universal for different downstream tasks. We provide comprehensive theoretical analyses of our method regarding its capability of handling node classification and graph classification as downstream tasks. Extensive experiments on ten graph datasets under four pre-training strategies demonstrate the superiority of our proposed method against six baselines. Our code is available at https://github.com/xbfu/EdgePrompt.

Problem

• Bridges gap between pre-training and downstream tasks
• Enhances graph representations by focusing on edges
• Improves adaptability of pre-trained GNN models

Innovation

• EdgePrompt manipulates input graphs via edge prompt vectors.
• Incorporates edge prompts through message passing in GNNs.
• Compatible with various GNN architectures and tasks.