The Taxonomy of Distributed Graph Analytics

Abstract

The rapid growth of large-scale networks such as Social networks, Heterogeneous and Complex networks generate huge volumes of network data. The analysis of networks has a significant role in obtaining node behavior relevant to different kinds of real-life applications. Graph analytics is an emerging area of graph-based data mining that describes the relationships among the nodes in a network. This paper aims to provide the taxonomy of various distributed programming models, distributed graph processing frameworks and various kinds of graph analytics that are essential for the analysis of large-scale networks. We categorize graph analytics tools into framework-based, cloud-based, Real-time based and In-memory based graph analytics tools. We compare and contrast distinct framework-based and cloud-based graph analytics tools with respect to several graph analytics features. Moreover, we present some of the open problems that are to be developed by distributed graph frameworks.