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.

How to use TraceClustering plugin export sublogs?

I have read the paper :Discovering Deviating Cases and Process Variants Using Trace Clustering,and I want to use it export the sublogs,but when I use it in prom,it only output a diagram that has a lot of circles.
Can anyone help?
Thank you in advance. 

Comments

  • Dear TangYahui,

    When you click on a cluster (represented by a circle), the right-hand parts of the screen should show you the option to export the cases in the selected cluster to the ProM workspace as a new event log. From there you can export it or use it as input for other plug-ins. You can also select multiple clusters by dragging an area or by holding the shift or ctrl keys.

    Kind regards,

    Bart Hompes
    Bart Hompes - Eindhoven University of Technology
  • Thank Bart,

    And it works.
    I notice that only choose picking mode can it works. This is why I could not export them before.
      
    And how can I choose the better "cluster dimensions"?

    Can prom secondary clustering data?Cause the result contains too many clusters.I want to reduce the number of clusters.




    Kind regards,
  • Hi TangYahui,

    Nice to hear you were able to solve the problem regarding exporting clusters as event logs.

    As for the question to improve your results, the answer is twofold:
    - What clustering dimensions you use determine what aspects will be considered to compute similarity between traces. As such, the things you want to cluster upon should be selected as dimensions.
    - To reduce the number of clusters you can play around with the expansion and inflation parameters of the MCL algorithm.

    You will find more information on both points in the paper and on related topics on this forum.

    Happy mining!

    Kind regards,

    Bart Hompes
    Bart Hompes - Eindhoven University of Technology
  • Hi Bart,

    Now I understand the concepts.But I noticed that when I deal with large files,the plug-in always become very slow.

    Can I use this plug-in generate fixed number of clusters?Through using the two parameters .

    Thank you again.

    Kind regards,

    Tang Yahui
Sign In or Register to comment.