arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2009.05303cs.LGstat.ML
收藏

CatGCN:具有范畴节点特征的图卷积网络

CatGCN: Graph Convolutional Networks with Categorical Node Features

University of Science and Technology of China(中国科学技术大学);National University of Singapore(新加坡国立大学);Google Research(谷歌研究院);Tencent WeChat(腾讯微信)

Weijian Chen, Fuli Feng, Qifan Wang, Xiangnan He, Chonggang Song, Guohui Ling, Yongdong Zhang

摘要:

最近对图卷积网络(GCN)的研究表明,初始节点表示(即,在第一次图形卷积之前的节点表示)很大程度上影响最终的模型性能。然而,当学习一个节点的初始表示,大多数现有的工作线性组合的节点特征的嵌入,而不考虑功能(或功能嵌入)之间的相互作用。我们认为,当节点特征是分类的,例如,在许多现实世界的应用中,如用户分析和推荐系统,特征交互通常携带用于预测分析的重要信号。忽略它们将导致次优的初始节点表示,从而削弱后续图卷积的有效性。在本文中,我们提出了一个新的GCN模型命名为CatGCN,这是专为图学习时的节点特征是分类的。具体来说,我们将两种显式交互建模方法集成到初始节点表示的学习中,即,在每对节点特征上进行局部交互建模,在人工特征图上进行全局交互建模。然后,我们使用基于邻域聚合的图卷积来细化增强的初始节点表示。我们以端到端的方式训练CatGCN,并在半监督节点分类上演示它。在腾讯和阿里巴巴数据集上进行的三个用户分析任务(用户年龄、城市和购买水平预测)的大量实验验证了CatGCN的有效性,特别是在图卷积之前执行特征交互建模的积极效果。

英文摘要:

Recent studies on Graph Convolutional Networks (GCNs) reveal that the initial node representations (i.e., the node representations before the first-time graph convolution) largely affect the final model performance. However, when learning the initial representation for a node, most existing work linearly combines the embeddings of node features, without considering the interactions among the features (or feature embeddings). We argue that when the node features are categorical, e.g., in many real-world applications like user profiling and recommender system, feature interactions usually carry important signals for predictive analytics. Ignoring them will result in suboptimal initial node representation and thus weaken the effectiveness of the follow-up graph convolution. In this paper, we propose a new GCN model named CatGCN, which is tailored for graph learning when the node features are categorical. Specifically, we integrate two ways of explicit interaction modeling into the learning of initial node representation, i.e., local interaction modeling on each pair of node features and global interaction modeling on an artificial feature graph. We then refine the enhanced initial node representations with the neighborhood aggregation-based graph convolution. We train CatGCN in an end-to-end fashion and demonstrate it on semi-supervised node classification. Extensive experiments on three tasks of user profiling (the prediction of user age, city, and purchase level) from Tencent and Alibaba datasets validate the effectiveness of CatGCN, especially the positive effect of performing feature interaction modeling before graph convolution.

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。

↑