Process mining discovers formal models that mimic business process dynamics by analyzing event logs containing execution traces, with classic algorithms extracting workflow models evaluated through conformance measures like fitness and precision. The stochastic extension incorporates trace frequencies to develop probabilistic models, typically by first discovering a standard workflow model and then optimizing parameters to match the observed frequency distribution in the log.

We present ProDiSt, a tool for PROcess DIscovery by STochastic approaches. The present version of the tool comes with three main functionalities:

Optimization of the parameters of a stochastic workflow net by exact computation of its stochastic language
Inference of the posterior distribution of the optimal parameters of a stochastic workflow net via a simulation-based Bayesian scheme
Optimization of the parameters of a stochastic process tree