arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2302.00890cs.LGcs.AIcs.SI
收藏

用于链接预测的带补全的神经共同邻居

Neural Common Neighbor with Completion for Link Prediction

Institute for Artificial Intelligence, Peking University(北京大学人工智能研究院);School of Intelligence Science and Technology, Peking University(北京大学智能科学与技术学院)

Xiyuan Wang, Haotong Yang, Muhan Zhang

AI总结:

本文提出NCN链接预测模型,利用结构特征指导MPNN池化,并针对图不完整性提出NCNC模型补全共同邻居结构,在标准基准上超越现有最先进模型。

摘要:

在这项工作中,我们提出了一种新颖的链接预测模型,并通过研究图的不完整性进一步提升了其性能。首先,我们引入了MPNN-then-SF,这是一种利用结构特征(SF)来指导MPNN表示池化的创新架构,其具体实现名为Neural Common Neighbor(NCN)。与现有模型相比,NCN展现出卓越的表达能力和可扩展性,现有模型可分为两类:SF-then-MPNN(用SF增强MPNN的输入)和SF-and-MPNN(将SF和MPNN解耦)。其次,我们研究了图的不完整性(即输入图中某些链接未被观测到的现象)对共同邻居等SF的影响。通过数据集可视化,我们观察到不完整性减少了共同邻居并引起分布偏移,显著影响模型性能。为解决这一问题,我们提出使用链接预测模型来补全共同邻居结构。将此方法与NCN结合,我们提出了带补全的神经共同邻居(NCNC)。NCN和NCNC大幅优于近期强大的基线模型,且NCNC在标准链接预测基准上进一步超越了最先进的模型。我们的代码可在https://github.com/GraphPKU/NeuralCommonNeighbor获取。

英文摘要:

In this work, we propose a novel link prediction model and further boost it by studying graph incompleteness. First, we introduce MPNN-then-SF, an innovative architecture leveraging structural feature (SF) to guide MPNN's representation pooling, with its implementation, namely Neural Common Neighbor (NCN). NCN exhibits superior expressiveness and scalability compared with existing models, which can be classified into two categories: SF-then-MPNN, augmenting MPNN's input with SF, and SF-and-MPNN, decoupling SF and MPNN. Second, we investigate the impact of graph incompleteness -- the phenomenon that some links are unobserved in the input graph -- on SF, like the common neighbor. Through dataset visualization, we observe that incompleteness reduces common neighbors and induces distribution shifts, significantly affecting model performance. To address this issue, we propose to use a link prediction model to complete the common neighbor structure. Combining this method with NCN, we propose Neural Common Neighbor with Completion (NCNC). NCN and NCNC outperform recent strong baselines by large margins, and NCNC further surpasses state-of-the-art models in standard link prediction benchmarks. Our code is available at https://github.com/GraphPKU/NeuralCommonNeighbor.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑