Graph Structure Learning (GSL) aims to capture intrinsic dependencies and interactions among nodes in graph-structured data by generating novel graph structures. Graph Neural Networks (GNNs) have emerged as promising solutions, but they often rely heavily on explicit graph structural information, making them susceptible to data noise and sparsity. To address these limitations, the authors propose GraphEdit, an approach that leverages large language models (LLMs) to learn complex node relationships in graph-structured data. By enhancing the reasoning capabilities of LLMs through instruction-tuning over graph structures, GraphEdit aims to overcome the limitations of explicit graph structural information and improve the reliability of graph structure learning. The model not only effectively denoises noisy connections but also identifies node-wise dependencies from a global perspective, providing a comprehensive understanding of the graph structure. Extensive experiments on multiple benchmark datasets demonstrate the effectiveness and robustness of GraphEdit across various settings. The model implementation is available at <https://github.com/HKUDS/GraphEdit>.Graph Structure Learning (GSL) aims to capture intrinsic dependencies and interactions among nodes in graph-structured data by generating novel graph structures. Graph Neural Networks (GNNs) have emerged as promising solutions, but they often rely heavily on explicit graph structural information, making them susceptible to data noise and sparsity. To address these limitations, the authors propose GraphEdit, an approach that leverages large language models (LLMs) to learn complex node relationships in graph-structured data. By enhancing the reasoning capabilities of LLMs through instruction-tuning over graph structures, GraphEdit aims to overcome the limitations of explicit graph structural information and improve the reliability of graph structure learning. The model not only effectively denoises noisy connections but also identifies node-wise dependencies from a global perspective, providing a comprehensive understanding of the graph structure. Extensive experiments on multiple benchmark datasets demonstrate the effectiveness and robustness of GraphEdit across various settings. The model implementation is available at <https://github.com/HKUDS/GraphEdit>.