Graph algorithms are very useful in solving many problems in all major domains like social networking, stock market analysis etc. Increasing demands of such kind of problems grow in scale and reveal the need of parallel computing resources to meet the computational and memory needs. Graph processing systems have to deal with the three Vs of big data - variety, velocity and volume. Loading the entire graph into the memory of a single machine seems to be impossible. In such cases, parallel processing is a solution to tackle the resource limitations posed by single processors. In this paper, we present the requirement for graph partitioning and the issues in designing partitioning technique for real world graphs. The paper also talks about the temporal metrics that provide a more effective analysis of real-world networks compared to their static counterparts. Finally, we aim at the reachability queries which are indispensable in networks and their usage in the dynamic graphs which evolve over time. We focus on the current challenges in this area and feature some future research recommendations. © Research India Publications.
cited By 0
R. G. Gayathri and Jyothisha J. Nair, “Towards efficient analysis of massive networks”, International Journal of Applied Engineering Research, vol. 10, pp. 222-227, 2015.