Monday August 31th, 2015, 16.00-18.00
University of Innsbruck, Universitätsstraße 15, 6020 Innsbruck, Austria

About 40 people joined the annual meeting. Moreover, 60 people joined the process mining reception organized by the taskforce on Sunday.

Slides of the meeting

Please check the action points. Being part of this task force requires active participation!

Short summary of the goals for the task force.

See the slides for the many activities that took place. Some highlights:

• another successful BPI workshop with record numbers of participants (65) and submissions,
• the BPI challenge providing datasets widely used by PM researchers and also used in education/training,
• the process mining MOOC with 68.000 registered participants,
• the LinkedIn group (number of members more than doubled over the last year),
• the 14th translation of the PM manifesto in Thai,
• the PM Camp with 170+ participants,
• the Best Process Mining Dissertation Award and publication in LNBIP,
• workshops and special sessions: ATAED 2015, CIDM 2014, WCCI 2014, CIDM 2015, etc.,
• special issue of IEEE TCS on process mining and big data,
• new PM movies, and
• new XES datasets.

The Best Process Mining Dissertation Award is a yearly award conferred by the IEEE Task Force on Process Mining to the author of an outstanding PhD thesis on business process intelligence. The winner of the second edition was announced during the task force's meeting. Dirk Fahland, Antonella Guzzo, and Marcello La Rosa organized the award that was handed out by Marcello to Jorge Munoz-Gama for this PhD thesis entitled “Conformance Checking and Diagnosis in Process Mining”. Jorge also gave a presentation summarizing his work. He received 1000 euro and will be invited to publish a book based on his thesis by Springer.

See slides for the update given by Felix. The formal processes are ongoing and, as required by the IEEE, a larger working group of 25+ persons has been established. All people accepted and support the standard. See https://standards.ieee.org/develop/project/1849.html on the IEEE SA website for more information. At the end of the year there will be a draft proposal that should be accepted at the end of 2016. The expectation is that the standards document will only require minor changes: The goal is to finalize the process and not start discussion on scope etc. These have been set quite some time ago. Mostly administrative hurdles need to be taken and improvements to the draft are solicited from all task force members (send input to Eric Verbeek). The full draft is available online shared:downloads:xes-r7.pdf. Anyone can comment and contribute to this effort.

[Action Point: All – Comment on draft before November 2015]

Moreover, for the community there are the following action points to make XES a widely used standard:

• More tools should support XES. Currently there are already 10 process mining tools supporting XES, but more tools could be added.
• Additional XES extensions should be developed using the extension mechanism, e.g., next to cost and alignments, also things like privacy, security, risks, reliability could be perspectives that can be added.
• More analysis techniques should exploit non-standard extensions.
• More datasets should be made available (contact Boudewijn van Dongen if not clear how to distribute via 3TU data center).

[Action Point: Vendor – Add XES support if not yet provided]
[Action Point: Researchers/Developers – Add extensions and analysis techniques exploiting these]
[Action Point: All – Provide data sets via 3TU data center]

Due to time constraints only few topics could be discussed in detail (see slides for other discussion topics). Main action points from the discussion: The working group on legal/ethical aspects of process mining by Frank van Geffen, Anne Rozinat, and Léonard Studer is ongoing and still considered relevant. However, at the same time there is a need for very practical guidelines for using/sharing event logs. This was discussed at the meeting and the following actions were proposed:

• Make a default or example NDAs. Not to go into legal or country specific issues, but to show the typical ingredients and some examples. This is also related to publications and anonymization.
• Define different levels of anonymization (e.g., a ranking system to taxonomy of sensitivity of logs) and possibly also confidence (to decide how reliable are conclusions derived from it can be).
• Guidelines to remove sensitive information.

If there would be a substantial effort on this, it could lead not only to information on the Task Force's website, but also to an LNBIP paper (like the manifesto). The following people volunteered to contribute to this effort:

• Andrea Burattin
• Jan Claes
• Felix Mannhardt

The new working group should contact Anne and Frank to check for overlap. The initiatives could also be merged, but this is left to the

[Action Point: Wil – Ensure the establishment of a working group]
[Action Point: working group (Felix Mannhardt, Andrea Burattin,Jan Claes) – Develop a document that can be discussed in the task force and that may lead to a publication]

Next to the yearly BPI challenge there is need for clear benchmark processes and logs. Focus is on discovery but could be extended. The BPI challenge considers a business setting where data needs to be interpreted. There could be a complimentary collection of benchmarks with quantifiable formally defined quality metrics. The following two persons will establish a new working group on this and involve others

• Massimiliano de Leoni
• Benoît Depaire

[Action Point: working group (Massimiliano de Leoni, Benoît Depaire) – Create a new benchmark within the next half year]

Other action points that resulted from the discussion:

[Action Point: All – Provide data sets via 3TU data center. For example the data used on BPI/BPM papers should be distributed via this route.]

[Action Point: All – Provide successful case studies for website (send to Wil).]

[Action Point: All – Find new ways to further promote process mining in your country. Also help to promote each other activities. For example the rerun of the Online Process Mining Course https://www.coursera.org/course/procmin starting October 7, 2015]

[Action Point: All Academics – Organize workshops & special sessions. Make sure to have a good distribution over the year, and aim for easy to reach locations.]

[Action Point: Akhil Kumar- Another attempt to have a process mining workshop track at one of the main data mining conferences.]

Many activities are already planned:

• BPI workshop 2016 in Rio,
• Rerun Online Process Mining Course https://www.coursera.org/course/procmin,
• BPI Challenge 2016 in Rio,
• Best Process Mining Dissertation Award 2016,
• Special issues, workshops, etc., e.g., ATAED 2016 in Toruń, Poland and possible PM sessions at WCCI 2016 (Vancouver), SSCI2016 (Greece), and main data mining conferences,
• Process Mining Camp 2016,
• New data sets, case studies, movies, …?,
• Additional Use Cases
• Additional Public Data Sets
• Additional MOOCs, books, etc. for end-users.

See also slides.

The meeting was followed by a reception at the Seegrube Alpenlounge at 1900 meter altitude.