C-Explorer: Browsing Communities in Large Graphs

Community retrieval (CR) algorithms, which enable the extraction of subgraphs from large social networks (e.g., Facebook and Twitter), have attracted tremendous interest. Various CR solutions, such as k-core and CODICIL, have been proposed to obtain graphs whose vertices are closely related. In this paper, we propose the C-Explorer system to assist users in extracting, visualizing, and analyzing communities. C-Explorer provides online and interactive CR facilities, allowing a user to view her interesting graphs, indicate her required vertex q, and display the communities to which q belongs. A seminal feature of C-Explorer is that it uses an attributed graph, whose vertices are associated with labels and keywords, and looks for an attributed community (or AC), whose vertices are structurally and semantically related. Moreover, C-Explorer implements several state-of-the-art CR algorithms, as well as functions for analyzing their effectiveness. We plan to make C-Explorer an open-source web-based platform, and design API functions for software developers to test their CR algorithms in our system.