Teoria espectral de grafos aplicada ao problema de isomorfismo de grafos

Abstract

In this work we investigated the use of concepts from Spectral Graph Theory (SGT) to support the construction of algorithms that solve the Graph Isomorphism Problem (GIP). Three theoretical results which consider information from the spectrum of the graphs and from the eigenvector centralities were presented. Furthermore, an algorithm for detection of graph isomorphism based on two of these results was proposed. Finally, we present the computational results comparing this algorithm with others from literature.