An Improved CRD (Close, Reach& Degree) algorithm to Process the Complex Graph Datasets

Authors

  • M. Sailaja, Dr. CVPR Prasad

Abstract

In data mining, graph mining is an important research area. This is also called as sub-domain and it focus on identifying the frequent sub-graphs within the graph datasets. Graph data is semi-structured and it is very complex data sets to process. While the difficulty of obtaining frequent sub-graphs in a large latent graph is severe, continuously finding frequent sub-graphs in graphs with streaming updates is harder. It is because the set of frequent sub-graphs has updated as every batch of node/edge update comes. In this paper, the new algorithm is implemented to find sub-graphs from large graph datasets using CRD (Close, Reach& Degree) algorithm.  To improve the performance of the CRD the Parallel Processing Algorithm (CRD-PPA) is adapted to overcome the issues like memory loading and computation time. CRD algorithm will find sub-graphs by evaluating distance and connectivity between vertex and running CRD algorithm on large dataset can reduce computation time.

Published

2020-10-16

Issue

Section

Articles