Title: Performance Analysis: Discovering Semi-Markov Models From Event Logs

Abstract: Process mining is a well-established discipline of data analysis focused on the discovery of process models from information systems' event logs. Recently, an emerging subarea of process mining - stochastic process discovery has started to evolve. Stochastic process discovery considers frequencies of events in the event data and allows for more comprehensive analysis. In particular, when durations of activities are presented in the event log, performance characteristics of the discovered stochastic models can be analyzed, e.g., the overall process execution time can be estimated. Existing performance analysis techniques usually discover stochastic process models from event data and then simulate these models to evaluate their execution times. These methods rely on empirical approaches. This paper proposes analytical techniques for performance analysis allowing for the derivation of statistical characteristics of the overall processes' execution times in the presence of arbitrary time distributions of events modeled by semi-Markov processes. The proposed methods can significantly simplify the what-if analysis of processes by providing solutions without resorting to simulation.