Process Discovery Contest 2024 has started
A new data set is now ready for discovery algorithms to be tested on. See the PDC 2024 page for details. Discovery algorithms can be submitted as of now.
Category: Process Discovery Contest
The Xcavation plugin scores well on the PDC 2023 data set
Using the default settings, the new Xcavation plugin scores 90.1% on the PDC 2023 data set, with a positive accuracy of 88.0% and a negative accuracy of 92.6%. Using the weights 50, 10 and 5 for fitness, precision and simplicity, it even scores 92.2%, with a positive accuracy of 89.9% and a negative accuracy of Read More …
PDC 2023 Data Set Disclosed
The data set of the Process Discovery Contest of 2023 (PDC 2023) has now been disclosed. It contains the following five folders (in ZIP archives): A trace from the Test log is classified against the corresponding trace from the Base log. If the trace from the Test log fits the model (strict) better than then Read More …
Release 6.13.105 of DiSCover
Changes the initial configuration of the Xcavation plugin to aconfiguration that provided good results on the PDC 2022 data set.
Discovering an S-Coverable WF-net using DiSCover
Presented at the ICPM 2022 International Conference on Process Mining in Bolzano, Italy, on October 25, 2022 Citation Slides Downloads Both algorithms use an absolute threshold that equals 1, a relative threshold that equals 3, and default values for all other parameters. These thresholds are set in the “Scripts/Discover.txt” file, lines 31 and 32. After Read More …
PDC 2022 Data Set Disclosed
The data set of the Process Discovery Contest of 2022 (PDC 2022) has now been disclosed. It contains the following five folders (in ZIP archives): A trace from the Test log is classified against the corresponding trace from the Base log. If the trace from the Test log fits the model (strict) better than then Read More …
New DiSCover algorithm performs well on PDC data sets
ProM 6.12 will contain a new discover algorithm called DiSCover. This algorithm discovers an accepting Petri net from an event log using a collection of DFGs (Directly Follows Graphs) and has been tested on the data sets of the Process Discovery Contests. As the DiSCover algorithm is non-deterministic, we ran it three times on every Read More …
Conversion of a log skeleton graph to a Petri net
Conversion For every activity in the log skeleton graph, a single transition is created. The fragments below show which places and additional transitions are required and how they get connected to the activity transitions. Attempts are made to not include a fragment if it will be subsumed by another fragment. Process The place {i} is Read More …
PDC 2019: Award
Yesterday, I received the “Process Discovery Contest Award” for my winning submission to the Process Discovery Contest 2019. Many thanks to Massimiliano, Benoit, and Josep for organizing the contest. Next year, I will help them :-).
195/200 Score for the Process Discovery Contest 2016
Using a fully-automated variant of the Log Skeleton Filter and Browser, I have managed to classify 195 out of 200 traces correctly for the Process Discovery Contest 2016. Not that this is better than my winning “DrFurby classifier” approach, which classified 193 out of 200 traces correctly. Below, you will find the main log skeletons Read More …