GO: A cluster algorithm for graph visualization. (June 2015)

• Record Type: Journal Article
• Authors: Huang, Xiaodi; Huang, Weidong
• Is Part Of: Journal of visual languages & computing. Volume 28(2015)
• Page Start: 71
• Page End: 82
• Publication Date: 2015-06
• Subjects: Information visualization -- Graph clustering; Visual programming languages (Computer science) -- Periodicals; Visual programming (Computer science) -- Periodicals; Programming languages (Electronic computers) -- Semantics -- Periodicals; Langages de programmation visuelle -- Périodiques; Programmation visuelle -- Périodiques; Langages de programmation -- Sémantique -- Périodiques; Programming languages (Electronic computers) -- Semantics; Visual programming (Computer science); Visual programming languages (Computer science); Periodicals; Electronic journals; 005
• DOI: 10.1016/j.jvlc.2014.12.007
• Languages: English
• ISSNs: 1045-926X

Abstract

Abstract: As we are in the big data age, graph data such as user networks in Facebook and Flickr becomes large. How to reduce the visual complexity of a graph layout is a challenging problem. Clustering graphs is regarded as one of effective ways to address this problem. Most of current graph visualization systems, however, directly use existing clustering algorithms that are not originally developed for the visualization purpose. For graph visualization, a clustering algorithm should meet specific requirements such as the sufficient size of clusters, and automatic determination of the number of clusters. After identifying the requirements of clustering graphs for visualization, in this paper we present a new clustering algorithm that is particularly designed for visualization so as to reduce the visual complexity of a layout, together with a strategy for improving the scalability of our algorithm. Experiments have demonstrated that our proposed algorithm is capable of detecting clusters in a way that is required in graph visualization. Highlights: We present an algorithm for clustering large graphs. The algorithm is able to detect the number of clusters automatically and the clusters with the suitable sizes in real- time from a given graph. The algorithm performs better than the compared algorithms against the real-world graphs.