Model conformance

The Model Conformance panel visualizes the fitness and conformance indicators:

• Similarity
  How the data-derived model compares with the reference model.
• Average fitness
  How the data-derived model is representative of the cases (some paths might be hidden, depending on the model details).
• Minimum fitness
  The least similar case to the data-derived model.
• Maximum fitness
  The most similar case to the data-derived model.

The fitness depends on the model details settings. The more the model details are increased, the more fitness there is (for example, if both activities and relations details are set to 100%, the fitness is 100%). By increasing the model details too much, the model becomes less readable.

The values are normalized from 0 (bad) to 1 (good).

With the Model Conformance feature, it is possible to do a visual conformance check between the Data-derived model and the Reference model. You can also click Compare to have a look at the similarities and the differences between the two of them.

Actions that are carried out in the Model Conformance box include:

• Data-derived model information
• Reference model information
• Compare section

Data-derived model information

By selecting Data-derived in the menu, IBM Process Mining visualizes the Data-Derived Model.

Reference model information

By selecting Reference in the menu, IBM Process Mining visualizes the Reference Model.

Compare section

By selecting Compare in the menu, IBM Process Mining automatically visualizes the resemblance between the two models.

The Comparison Legend (at the lower right of the page)

• Red box or arrow indicate that the activity or transition are only present in the Derived model
• Yellow box or arrow indicate that the activity or transition are only present in the Reference model
• Blue box (dark or light blue depending on the frequency) indicates that the activity is present in both models
• Black arrow indicates that the transition is present in both models