Computing the extent of conformance of a given trace log and a petri-net model

ashishashish Posts: 4
edited September 2014 in ProM 6

I am trying to compute the extent of conformance of a given trace log and a petri-net model. I used “Replay a Log on Perti Net for Conformance”. The output gives 3 metrics on conformance: trace fitness, move-model fitness and move-log fitness.

(Referring to a paper) Trace fitness value represent the fitness value of the Petri Net with the log. This value indicates how well the event log can be replayed in the discovered Petri Net. A fitness value of ‘1’ means that the log can be successfully replayed, whereas a value of ‘0’ means that this is completely not the case.


What is the meaning of move-model fitness and move-log fitness? Can some please explain in simple language rather than referring to a paper which I have gone through but I am not able to fully understand it?

Also, I did not see appropriateness metrics (simple behavioral appropriateness and simple structural appropriateness).

Post edited by ashish on


  • JBuijsJBuijs Posts: 922

    Deas Ashish,

    When an event log is replayed on a Petri net 3 types of moves can be made:

    - synchronous: both an activity in the process model (/Petri net) as well as an event in the current trace can be 'moved', i.e. there is no issue or misalignment here.

    - Model move only: there is a mismatch between the observed activities in the traces and the possible activities in the process model. To get an aligment with the fewest mismatches, the algorithm decides that the model will move one activity forward while the trace remains as is. This means that according to the model an activity was expected but not observed.

    - Log move only: this is the contrary of the situation discussed above. Here the algorithm makes the trace move a step while the model remains in the current position.

    Hope this helps in your further understanding of alignments.

    To answer your second question: as far as I know these are not implemented as seperate metrics in ProM 6, you might find them in ProM 5.2

    Joos Buijs

    Senior Data Scientist and process mining expert at APG (Dutch pension fund executor).
    Previously Assistant Professor in Process Mining at Eindhoven University of Technology
Sign In or Register to comment.