An edge-based framework for fast subgraph matching in a large graph

Abstract

In subgraph matching, we want to find all subgraphs of a database graph that are isomorphic to a query graph. Subgraph matching requires subgraph isomorphism testing, which is NP-complete. Recently, some techniques specifically designed for subgraph matching in a large graph have been proposed. They are based on a filtering-and-verification framework. In the filtering phase, they filter out vertices that are not qualified for subgraph isomorphism testing. In the verification phase, subgraph isomorphism testing is performed and all matched subgraphs are returned to the user. We call them a vertex-based framework in the sense that they use vertex information when filtering out unqualified vertices. Edge information, however, can also be used for efficient subgraph matching. In this paper, we propose an edge-based framework for fast subgraph matching in a large graph. By using edge connectivity information, our framework not only filters out more vertices in the filtering phase, but also avoids unnecessary edge connectivity checking operations in the verification phase. The experimental results show that our method significantly outperforms existing approaches for subgraph matching in a large graph.