Diversified Pattern Mining on Large Graphs

Xin Wang, Liang Tang, Yong Liu, Huayi Zhan, Xuanzhe Feng

Abstract: Frequent pattern mining ( $$\mathsf {FPM}$$ ) on large graph has been receiving increasing attention due to its wide applications. The $$\mathsf {FPM}$$ problem is defined as mining all the subgraphs (a.k.a. patterns), with frequency above a user-defined threshold in a large graph. Though a host of techniques have been developed, most of them suffers from high computational cost and inconvenient result inspection. To tackle the issues, we propose an approach to discover diversified top-k patterns from a large graph G. We formalize the distributed top-k pattern mining problem based on a diversification function. We develop an algorithm with early termination property, to efficiently identify diversified top-k patterns. Using real-life and synthetic graphs, we show advantages of our algorithm via intensive experimental studies.