Matching Graphs with Unique Node Labels

Abstract

In its most general form, graph matching refers to the problem of finding a mapping f from the nodes of one given graph g₁ to the nodes of another given graph g₂ that satisfies some constraints or optimality criteria. For example, in graph isomorphism detection [130], mapping f is a bijection that preserves all edges and labels. In subgraph isomorphism detection [173], mapping f has to be injective such that all edges of g₁ are included in g₂ and all labels are preserved. Other graph matching problems that require the constructions of a mapping f with particular properties are maximum common subgraph detection [118, 129] and graph edit distance computation [131, 151].