ABSTRACT
Graphs are a powerful and versatile tool useful in various subfields of science and engineering. In recent years, the use of graph representations has gained popularity in pattern recognition and machine learning [1-3].The main advantage of graphs is that they offer a powerful way to represent structured data.
The process of evaluating the structural similarity of graphs is referred to as graph matching. A large variety of methods addressing specific problems of structural matching have been proposed [13].
In this paper we provide an introduction to, and overview of, basic graph matching methods. We describe exact matching paradigms, such as graph isomorphism, subgraph isomorphism and maximum common subgraphs as well as error- tolerant matching approaches, such as graph edit distance.
Keywords: Graphs, Graph matching, Exact graph matching, Error-tolerant graph matching, Graph isomorphism, Subgraph isomorphism, Maximum common subgraph, Error- tolerant matching.