Simulation-Based Graph Similarity

We present symmetric and asymmetric similarity measures for labeled directed rooted graphs that are inspired by the simulation and bisimulation relations on labeled transition systems. Computation of the similarity measures has close connections to discounted Markov decision processes in the asymmetric case and to perfect-information stochastic games in the symmetric case. For the symmetric case, we also give a polynomial-time algorithm that approximates the similarity to any desired precision. Comments Postprint version. Published in Lecture Notes in Computer Science, Volume 3920, Tools and Algorithms for the Construction and Analysis of Systems: Proceedings of 12th International Conference on Tools and Algorithms for the Construction and Analysis of Systems(TACAS 2006), pages 426-440. Publisher URL: http://dx.doi.org/10.1007/11691372_28 This conference paper is available at ScholarlyCommons: http://repository.upenn.edu/cis_papers/237 Simulation-Based Graph Similarity? Oleg Sokolsky, Sampath Kannan, and Insup Lee Department of Computer and Information Science University of Pennsylvania {sokolsky,kannan,lee}@cis.upenn.edu Abstract. We present symmetric and asymmetric similarity measures for labeled directed rooted graphs that are inspired by the simulation and bisimulation relations on labeled transition systems. Computation of the similarity measures has close connections to discounted Markov decision processes in the asymmetric case and to perfect-information stochastic games in the symmetric case. For the symmetric case, we also give a polynomial-time algorithm that approximates the similarity to any desired precision. We present symmetric and asymmetric similarity measures for labeled directed rooted graphs that are inspired by the simulation and bisimulation relations on labeled transition systems. Computation of the similarity measures has close connections to discounted Markov decision processes in the asymmetric case and to perfect-information stochastic games in the symmetric case. For the symmetric case, we also give a polynomial-time algorithm that approximates the similarity to any desired precision.