University of Twente Student Theses


Quantity meets quality for mined process models through simulated events

Arends, Simon Daniël (2020) Quantity meets quality for mined process models through simulated events.

[img] PDF
Abstract:Real-life systems and simulated models often generate event logs of their processes. These event logs can be mined to create process models that represent and visualize what happened throughout the process. Simulation models are particularly useful since they mimic the real world, with the benefit that it is possible to examine many different variations. To conduct simulation experiments, such models often have to run for a certain amount of time to generate event logs. This research investigates what the effect is of the number of event logs generated by such a simulation model, on the quality of the resulting mined process models. The research partitions a data set into subsets that simulate an increasing quantity of event logs. These subsets are then investigated using performance and conformance analysis. It is found that as the number of event logs increases, the analysis output becomes more reliable and the performance and conformance values stabilize and converge to certain values. A benefit of the approach is that it can lead to a method to assess whether a sufficient amount of event logs is generated, to mine representative process models. This can lead to saving time and more reliable results.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science, 58 process technology, 85 business administration, organizational science
Programme:Business & IT BSc (56066)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page