Abstract

Among various available graph patterns, one can discover the frequent substructures from a set of graphs. They are useful for characterizing graph sets, finding difference among groups of graphs, classifying and clustering graphs, and building graph indices leading to knowledge extraction. The Apriori-based approach, one such technique uses the breadth-first search (BFS) strategy because of its level-wise candidate generation. In this paper, we have started our work with representing the graphs, studying pattern growth and analysing the existing algorithm. Further, we have proposed our derived technique in the same direction. Our proposed technique is observed to be slightly efficient as compared to the existing algorithm. The paper concludes with analysis on the proposed technique with supportive examples, output, results and screen shot.