Community-enhanced Network Representation Learning for Network Analysis.

Cunchao Tu,Hao Wang,Xiangkai Zeng,Zhiyuan Liu,Maosong Sun

2016-01-01

Abstract:Network representation learning (NRL) aims to build low-dimensional vectors for vertices in a network. Most existing NRL methods focus on learning representations from local context of vertices (such as their neighbors). Nevertheless, vertices in many complex networks also exhibit significant global patterns widely known as communities. It's a common sense that vertices in the same community tend to connect densely, and usually share common attributes. These patterns are expected to improve NRL and benefit relevant evaluation tasks, such as link prediction and vertex classification. In this work, we propose a novel NRL model by introducing community information of vertices to learn more discriminative network representations, named as Community-enhanced Network Representation Learning (CNRL). CNRL simultaneously detects community distribution of each vertex and learns embeddings of both vertices and communities. In this way, we can obtain more informative representation of a vertex accompanying with its community information. In experiments, we evaluate the proposed CNRL model on vertex classification, link prediction, and community detection using several real-world datasets. The results demonstrate that CNRL significantly and consistently outperforms other state-of-the-art methods. Meanwhile, the detected meaningful communities verify our assumptions on the correlations among vertices, sequences, and communities.