Answering Graph Pattern Matching Using Views: A Revisit

Abstract

This paper studies how to answer graph pattern matching defined in terms of subgraph isomorphism by using a set of materialized views. We first propose a notion of pattern containment to characterize graph pattern matching using graph pattern views, and show that graph pattern matching can be answered using a set of views if and only if the pattern query is contained by the views, and develop efficient algorithm to determine pattern containment. Based on this characterization, an efficient algorithm is developed to evaluate graph pattern matching using views. In addition, when a pattern query is not contained in a set of views, we study the problem of approximately answering graph pattern matching using views. We first study maximally contained (resp. containing) rewriting problems, develop algorithms to find such rewritings. We then propose techniques to find approximate answers using maximally contained (resp. containing) rewriting. Using real-life and synthetic data, we experimentally verify that these methods are able to efficiently conduct graph pattern matching on large social graphs.