Conformance Metrics

Conformance Checking with mpmX specifies a process model (one or more happy paths)) to be calculated. Each process path in the log has the most similar happy path assigned to it. Then the sequence of activities of each real process variant is compared with the sequence of the happy path.

Activities and edges can have the following properties:

• They are correct, which means the observed behavior corresponds to the behavior in the process model.
• They are deviating, which means the observed behavior cannot be found in the process model.
• They were skipped, which means they are listed in the process model, but do not occur in the real process path that was followed.

The mpmX Conformance Checking algorithm evaluates the characteristics of each edge and node and calculates the Process Conformance metric.

• This metric shows the deviation between the model and reality in a range of 0-100%.
• This allows the analyst to quickly identify particularly good or bad process variants on the Conformance sheet and also to clearly see which deviation has occurred.