This repository curates research papers on prompt learning for Graph Neural Networks (GNNs), addressing the limitations of traditional "pre-train, fine-tune" workflows. It serves as a comprehensive resource for researchers and practitioners exploring the "pre-train, prompt" paradigm in graph-based machine learning.

How It Works

The repository categorizes papers based on their approach to prompt learning on graphs, including GNN prompting, multi-modal integration, and domain adaptation. It highlights how prompt learning, inspired by NLP, offers a more efficient and effective alternative for diverse downstream tasks on graph data. The core idea is to adapt pre-trained GNNs to specific tasks by learning task-specific prompts, rather than full fine-tuning.

Quick Start & Requirements

This repository is a curated list of papers and does not involve direct code execution or installation. It links to relevant code repositories for individual papers.

Highlighted Details

• Comprehensive coverage of prompt learning techniques for GNNs, including theoretical analyses, multi-modal applications, and domain adaptation strategies.
• Links to numerous open-source code implementations for many of the surveyed papers.
• Includes a dedicated benchmark suite (ProG) for evaluating graph prompt learning methods across various datasets and tasks.
• Provides a detailed table of commonly used datasets in GNN prompting research.

Maintenance & Community

The repository is actively maintained, with a call for contributions via issues or pull requests for new papers or corrections. It is based on a survey paper and aims for frequent updates.

Licensing & Compatibility

The repository itself is not licensed for software use. Individual linked papers and code repositories will have their own licenses.

Limitations & Caveats

This is a curated list of research papers, not a software library. Users must refer to individual paper repositories for code, dependencies, and execution instructions. The rapid pace of research means the list may not be exhaustive at any given moment.