Token-Based Replay

Token-based replay is a conformance-checking algorithm used in process mining that evaluates how well an actual recorded sequence of events, or trace, matches a process model by replaying that trace on a Petri net representation of the model. As the trace is replayed, the algorithm tracks four counts, the tokens produced, consumed, missing and left over, and uses them to compute a fitness value that expresses how closely the observed behavior conforms to the modeled process. This description is adapted from Wikipedia contributors under CC BY-SA 4.0; changes were made. https://creativecommons.org/licenses/by-sa/4.0/