An efficient algorithm for finding frequent and diverse subgraph patterns in PPI networks

Abstract

Finding frequent subgraph patterns in uncertain graph databases become an important problem with applications. The main difficulty in solving the problem is finding frequent subgraph patterns from the large number of candidate graph and testing large number of subgraph isomorphism from the graphs that contain a given pattern. The traditional method of finding frequent subgraph patterns cannot distinguish the diverse patterns of uncertain graph database. In this paper, we propose a method named mining uncertain and diverse subgraph pattern (MUDSP) in PPI networks. It is used to find the different subgraph pattern including high proportion or high probability pattern. The algorithm is also enables necessary optimisations with respect to pruning index tree. The evaluation of our approach on real-world datasets of uncertain graph databases demonstrates the efficiency with respect to approach of finding diverse pattern.