To prevent spam users, you can only post on this forum after registration, which is by invitation. If you want to post on the forum, please send me a mail (h DOT m DOT w DOT verbeek AT tue DOT nl) and I'll send you an invitation in return for an account.

Multitasking process

My data has this format:

Imgenes integradas 1

Considering the data (highlighted) having the same timestamp, would be possible to find as final answer the following petri-net attached in ProM?

Imgenes integradas 2

Are there some version of the ProM thats deals with multitasking? Are there some issue that I am skipping?

Answers

  • JBuijsJBuijs Posts: 910
    Dear Haydee,

    Could you please include the pictures? Uploading does not work properly, so I suggest using imgur.com or tinypic.com to upload and link.
    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
  • JBuijsJBuijs Posts: 910
    Dear Haydee,

    (first of all: good to see that forum uploads still work, I had issues earlier)

    Well, the assumption of most algorithms is that the events within a trace have an absolute order (e.g. strictly follow each other). start and complete information would be required for some algorithms to be able to detect this, although most wouldn't.
    you would need several traces where the activities appear in different orders for the model to be discovered by process discovery algorithms.
    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
  • HaydeeHaydee Posts: 7
    edited November 2016
    Thanks for your reply. But it does not solve my problem I know that making permutations and thus having different orders solves the parallelism but I have records with 140 activities and to make a permutacion would leave the xes file with too much information and could not take into account frequencies nor anything because the information Is not true
    Post edited by Haydee on
  • JBuijsJBuijs Posts: 910
    There is not an easy answer to your question. Each discovery algorithm has a different threshold when they interpret the activities as being executed in parallel. You don't need all possible combinations in your event log, just 'enough'.

    BTW: your Petri net contains a deadlock since activity F cannot fire if there is not token in the place from G to F.
    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.