JPMiner: Mining Frequent Jump Patterns from Graph Databases

A major challenge in frequent subgraph mining is the sheer size of its mining results. In many cases, allow minimum support may generate an explosive number of frequent subgraphs, which severely restricts the usage of frequent sub graph mining. In this paper, we study anew problem of mining frequent jump patterns from graph databases. Mining frequent jump patterns can dramatically reduce the number of output graph patterns, and still capture interesting graph patterns. By integrating the operation of checking jump patterns into the well-known DFS code tree enumeration framework, we present an efficient algorithm JPMiner for this new problem. We experimentally evaluate various aspects of Jupiter using both real and synthetic datasets. Experimental results demonstrate that the number of frequent jump patterns is much smaller than that of closed frequent graph patterns, and JPMiner is efficient and scalable in mining frequent jump patterns.