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.

Mining subprocesses

JBuijsJBuijs Posts: 912
edited December 2010 in ProM 6
Pieter van Gorp asked the following questions on the ProM mailing lists on December 13:

Hi all,
can somebody please provide a state-of-the-art w.r.t. process mining with support for subprocesses?

A quick google learned me that subprocesses are mentioned in http://prom.win.tue.nl/research/wiki/_media/publications/jcbose2009b.pdf
but I wonder which mining plugins support the mining of subprocess definitions ans calls.  I also wonder in which documented case studies such plugins have been evaluated so far.

Sincerely,
Pieter Van Gorp
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

Comments

  • JBuijsJBuijs Posts: 912
    To which JC replied:

    Dear Pieter,

    The most recent work that enables the mining of hierarchical process models (process maps) is the one of http://www.win.tue.nl/~jcbose/Li_JCBose_Aalst_BPI2010_ExtendedAbstract.pdf.
    This is a follow up to the abstractions paper that you referred to. You can find one case study in this paper. The discovery of process maps has been enabled/realized in ProM 6.0. In order to discover process maps, you need to use two packages in ProM 6.0 in the following order:

    (i) Pattern Abstractions (This is implemented as a Visualizer). This plugin implements the functionality of (Abstractions in Process Mining: A Taxonomy of Patterns) and also provides rich functionality to prune/select patterns, define abstractions (sub-processes) based on the common execution patterns, and transform the log (as defined in the latest paper mentioned above). An analyst need to explore the patterns, select/define meaningful abstractions and use the transformation functionality. The transformation creates sub-logs for each of the sub-processed defined.
    (ii) Fuzzy Miner. The original fuzzy miner has been enhanced to utilize the availabilty of sub-logs and mine sub-processes from these sub-logs. Just like cluster nodes, we now have abstract nodes in the fuzzy model which upon zooming-in can provide the sub-process captured underneath it.

    Note that Fuzzy Miner is just one realization of step 1 above. You can customize any other process discovery algorithm to make use of the sub-logs and mine process maps.

    We have recently applied this two-phase approach to process map discovery in the context of discovering the workflow of field service engineer at Philips the results of which is not yet publicly available. However, I would like to mention that this approach showed good promise in discovering comprehensible process maps.

    Hope this helps. Should you have any queries, please feel free to contact me.

    Thanks and Kind Regards,
    JC

    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: 912
    To which Pieter replied:

    Thanks for your very informative reply.  Ton Weijters will also send me a recently completed TUE master thesis on the topic.

    Regards,
    Pieter Van Gorp, Assistant Professor (UD)
        Information Systems Group, School of Industrial Engineering
        Eindhoven University of Technology (TU/e), The Netherlands
        Phone: +31 40 247 2062, Skype: pvgorp, Fax: +31 40 243 2612
        http://is.tm.tue.nl/staff/pvgorp/research/

    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.