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.

Merge traces

rbutler
edited June 2011 in - Usage
Hi there,

I have several self-contained structured traces and I want to merge them into one model. Does ProM provide a PlugIn to do so?

Thanks,
Ralf

Comments

  • Hi Ralf,

    First of all, welcome to the forum!

    I'm assuming that you have several logs (/sets) of traces and you want to combine them in one bigger log (/set).
    The ProM 6 development version has a plug-in that can do that called 'Log Merge' by Jan Claes.
    If you download the nightly build of ProM 6 and start the Package Manager you should be able to install the package.

    Let me know if this solves your problem!
    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
  • rbutler
    edited June 2011
    Hi Joos,

    thanks for the warm welcome ;)

    Sorry I was not very precise with my questions. It is not about merging logs. I have a xes file already. However the traces in there (roughly 50) need to be merged into one process model. I now saw that this is doable with the fuzzy miner or alpha algorithm. But the results in my case look horrible. Well, that is probably more related to the arbitrary execution of events and less to the miner plugins.
    In my case not a single event is repeated within the same trace. It would be nice to see a process model which combines all of those traces but keeps the amount of loops to a minimum. That would mean that couple of events would be shown multiple times within the process model but it would at least increase readability (at least I hope so ;)

    Thanks,
    Ralf

    Btw. is there still the Multi-Phase Miner in ProM6?
  • Hi Ralf,

    All process discovery algorithms produce 1 process model on which all/most of the trace can be replayed.
    Have you tried the (flexible) heuristic miner? In general it produces less spaghetti like models than the alpha miner. Furthermore, in the Fuzzy miner you can control the process model output using the sliders on the right.

    As far as I know, no process discovery algorithm is able to split activities to reduce the process model complexity.

    You can easily convert an XES event log to MXML using ProM 6. When you loaded your XES log in ProM 6, click on the 'export' button on the bottom right and choose to save it as (compressed) MXML. That way, you can also use ProM 5.2 for your analysis.
    At the moment the Multi-Phase miner is not available in ProM 6. But you're more than welcome to implement it ;)
    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
  • Thanks for this comprehensive answer Joos!
Sign In or Register to comment.