Multi-perspective Process Mining

I defended my PhD thesis Multi-perspective Process Mining on the 7th February 2018. The thesis provides in-depth description of several Process Mining techniques that deal with situations in which multiple process perspectives (control-flow, data, resource, time, etc.) are intertwined and considered together.

PhD thesis -- Multi-perspective Process Mining

Extended Abstract

If you just want to get an overview, an extended abstract summarizing it is published in the BPM Dissertation Award proceedings, for which I was nominated.

Technical Content

Of course, I am biased, but I think that despite its length (425 pages) it serves as nice introduction to Multi-perspective Conformance Checking (Chapter 4 and 5). Of course there are much more topics covered, such as Chapter 8 on Data-aware Heuristics Process Discovery in which the well-known Heuristics Miner algorithm is extended with an approach to detect data-dependent behavior and Chapter 9 on Guided Process Discovery in which existing process discovery methods can be improved by using multi-perspective activity patterns to raise the abstraction level of the event log. The thesis chapters are fairly independent from each other.

Tools

I like to actually realize my ideas as working software. I take pride in going the extra mile to make software that is actually usable and can be built upon by other people.

Multi-perspective Process Explorer (MPE)

TODO

Interactive Data-aware Heuristic Miner (iDHM)

TODO

XESLite

TODO

Case Studies

As part of my PhD I applied my technique to real processes. Four of those applications are described in the thesis showing that the application of methods is feasible and provides valuable insights. In three cases I managed to publish the underlying anonymized event data contributing to the growing set of realistic data available at the IEEE Task Force on Process Mining repository.

Many others have built their work upon these logs since then: 17 citations for the road traffic fine management log and 10 citations for the sepsis log so far with a high number of cases in which the data was not properly cited. You are welcome and please start citing data properly!

Road traffic fine management

TODO

Sepsis cases

TODO

Digital whiteboard

TODO

Hospital billing

TODO