期刊
ADVANCES IN KNOWLEDGE DISCOVERY AND DATA MINING, PAKDD 2021, PT I
卷 12712, 期 -, 页码 541-553出版社
SPRINGER INTERNATIONAL PUBLISHING AG
DOI: 10.1007/978-3-030-75762-5_43
关键词
-
类别
资金
- NSF [1447788, 1704074, 1757916, 1834251]
- Army Research Office [W911NF1810344]
- Intel Corp
- Digital Technology Center at the University of Minnesota
This paper introduces a new method for node representation learning that leverages coarse-grain information for clustering and maximizing mutual information among nodes, showing superior performance in tasks such as node classification, link prediction, and node clustering compared to competing methods.
This work proposes a new unsupervised (or self-supervised) node representation learning method that aims to leverage the coarse-grain information that is available in most graphs. This extends previous attempts that only leverage fine-grain information (similarities within local neighborhoods) or global graph information (similarities across all nodes). Intuitively, the proposed method identifies nodes that belong to the same clusters and maximizes their mutual information. Thus, coarse-grain (cluster-level) similarities that are shared between nodes are preserved in their representations. The core components of the proposed method are (i) a jointly optimized clustering of nodes during learning and (ii) an Infomax objective term that preserves the mutual information among nodes of the same clusters. Our method is able to outperform competing state-of-art methods in various downstream tasks, such as node classification, link prediction, and node clustering. Experiments show that the average gain is between 0.2% and 6.1%, over the best competing approach, over all tasks. Our code is publicly available at: https://github.com/cmavro/Graph-InfoClust-GIC.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据