Query-Aware Learnable Graph Pooling Tokens as Prompt for Large Language Models
This work addresses the problem of efficiently processing textual-attributed graphs for domains like social networks and knowledge graphs, representing an incremental advancement in graph-LLM integration.
The paper tackles the challenge of integrating large language models with graph-structured data by proposing Learnable Graph Pooling Token (LGPT) and Early Query Fusion to address scalability and information loss issues, achieving a 4.13% performance improvement on the GraphQA benchmark without retraining the LLM.
Graph-structured data plays a vital role in numerous domains, such as social networks, citation networks, commonsense reasoning graphs and knowledge graphs. While graph neural networks have been employed for graph processing, recent advancements have explored integrating large language models for graph-based tasks. In this paper, we propose a novel approach named Learnable Graph Pooling Token (LGPT), which addresses the limitations of the scalability issues in node-level projection and information loss in graph-level projection. LGPT enables flexible and efficient graph representation by introducing learnable parameters that act as tokens in large language models, balancing fine-grained and global graph information. Additionally, we investigate an Early Query Fusion technique, which fuses query context before constructing the graph representation, leading to more effective graph embeddings. Our method achieves a 4.13\% performance improvement on the GraphQA benchmark without training the large language model, demonstrating significant gains in handling complex textual-attributed graph data.