Title: Community detection for network with high dimensional attributes
Authors: Wanjie Wang - National University of Singapore (Singapore) [presenting]
Abstract: Community detection in social network is a topic with much interest nowadays due to the high demand. With the observed connections between nodes, it is of interest to cluster the nodes into different communities, so that nodes within the same community have larger probability to connect. With the development of technology, the observed data are not only the connections between nodes, but also the attributes of each node. For example, the abstract of each paper in the paper citation network. With the attributes, some works present new methods for a better community detection result. However, most of the works are on low-dimensional attributes, while now the attributes are quite high-dimensional. We propose a new algorithm for the community detection problem for social network with high dimensional attributes, with good theoretical properties and simulation results.