让图神经网络学会看图，提升链接预测效果
Open Your Eyes: Vision Enhances Message Passing Neural Networks in Link Prediction
- 引入视觉感知机制，增强图神经网络对结构的理解
- 在7个数据集上均实现性能提升，部分达新最优
- 适合关注图学习与多模态融合的研究者
消息传递图神经网络（MPNNs）和结构特征（SFs）是链接预测任务的核心。然而，作为普遍且直观的理解方式，视觉感知在MPNN领域尚未被充分挖掘。本文首次提出图视觉网络（GVN），通过引入有效的视觉结构感知机制，显著提升MPNN的表征能力，并推出更高效的变体E-GVN。大量实验证明，所提框架在7个链接预测数据集上均取得一致性能提升，包括大规模复杂图。该改进与现有先进方法兼容，且多个场景下达到新的最优结果，揭示了链接预测中视觉增强的新方向。
原文摘要 · Abstract (English)
Message-passing graph neural networks (MPNNs) and structural features (SFs) are cornerstones for the link prediction task. However, as a common and intuitive mode of understanding, the potential of visual perception has been overlooked in the MPNN community. For the first time, we equip MPNNs with vision structural awareness by proposing an effective framework called Graph Vision Network (GVN), along with a more efficient variant (E-GVN). Extensive empirical results demonstrate that with the proposed frameworks, GVN consistently benefits from the vision enhancement across seven link prediction datasets, including challenging large-scale graphs. Such improvements are compatible with existing state-of-the-art (SOTA) methods and GVNs achieve new SOTA results, thereby underscoring a promising novel direction for link prediction.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。