Search papers, labs, and topics across Lattice.
This paper introduces PromptGNN-sim, a bi-directional framework that integrates Graph Neural Networks (GNNs) and Large Language Models (LLMs) for enhanced learning on Text-Attributed Graphs (TAGs). By employing a Graph Attention Network (GAT) for semantically aware neighborhood selection and generating structure-aware prompts for LLMs, the method facilitates deeper interaction between textual semantics and graph structure. Experimental results across six datasets, including Cora and Pubmed, reveal that PromptGNN-sim significantly outperforms traditional GNNs, LLMs, and existing fusion techniques, highlighting the benefits of collaborative learning in graph contexts.
Interactive structure-semantic collaboration in PromptGNN-sim leads to superior performance in text-attributed graph learning, outperforming both classical and recent fusion methods.
Text-Attributed Graphs (TAGs) combine textual semantics with graph structure and are central to many graph learning tasks. However, existing fusion methods often treat text and structure as separate inputs in a shallow, one-way pipeline, which limits deep interaction between modalities and weakens performance under sparse connectivity or cross-graph generalisation. To address this issue, we propose PromptGNN-sim, a bi-directional structure-semantic fusion framework for collaborative GNN-LLM learning. PromptGNN-sim uses a Graph Attention Network (GAT) for semantically aware neighborhood selection by combining structural attention with textual similarity. The selected structural context is then used to generate structure-aware prompts for an LLM, including the target node summary, label categories, and representative keywords from similar neighbors. During training, bi-directional cross-modal contrastive learning and cross-attention are introduced to jointly optimize the GNN and LLM components. Experiments on six public datasets, including Cora, Pubmed, and WikiCS, evaluate accuracy, generalisation, and robustness under cross-task transfer, cross-dataset generalisation, and sparse perturbations. Results show that PromptGNN-sim outperforms classical GNNs, LLMs, and recent GNN-LLM fusion methods, demonstrating the effectiveness of interactive structure-semantic collaboration for text-attributed graph learning.
