Unified Graph Prompt Learning via Low-Rank Graph Message Prompting

2026-04-13

AI Summary

This work addresses the limitation of existing graph prompting methods, which are typically designed for individual graph components—such as node or edge features—and lack a unified framework. To overcome this, the authors propose a novel Graph Message Prompting (GMP) paradigm that enables joint prompt learning across all graph components for the first time. Building upon GMP, they further introduce Low-Rank Graph Message Prompting (LR-GMP), which incorporates low-rank structural constraints to enhance both prompting efficiency and model generalization while improving robustness. Leveraging graph neural networks and message-passing mechanisms, LR-GMP demonstrates significant performance gains over current state-of-the-art methods across multiple benchmark datasets, confirming its effectiveness and superiority in diverse downstream tasks.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Natural Language Processing: Prompt Engineering / Prompting
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Graph Data Prompt (GDP), which introduces specific prompts in graph data for efficiently adapting pre-trained GNNs, has become a mainstream approach to graph fine-tuning learning problem. However, existing GDPs have been respectively designed for distinct graph component (e.g., node features, edge features, edge weights) and thus operate within limited prompt spaces for graph data. To the best of our knowledge, it still lacks a unified prompter suitable for targeting all graph components simultaneously. To address this challenge, in this paper, we first propose to reinterpret a wide range of existing GDPs from an aspect of Graph Message Prompt (GMP) paradigm. Based on GMP, we then introduce a novel graph prompt learning approach, termed Low-Rank GMP (LR-GMP), which leverages low-rank prompt representation to achieve an effective and compact graph prompt learning. Unlike traditional GDPs that target distinct graph components separately, LR-GMP concurrently performs prompting on all graph components in a unified manner, thereby achieving significantly superior generalization and robustness on diverse downstream tasks. Extensive experiments on several graph benchmark datasets demonstrate the effectiveness and advantages of our proposed LR-GMP.

Problem

• Graph Prompt Learning
• Unified Prompting
• Graph Data Prompt
• Graph Components
• Low-Rank Representation

Innovation

• Graph Prompt Learning
• Low-Rank Representation
• Unified Prompting
• Graph Message Passing
• Pre-trained GNNs