Bachelor Thesis - Discovering High Quality Process Models Efficiently: Heuristic Approaches for Selecting Interesting Places

Contact

Name

Lisa Mannel

Scientific Assistant

Phone

work
+49 241 80 21916

Email

E-Mail
 

Description

More and more processes executed in companies are supported by information systems, which store each event executed in a so-called event log. In the context of process mining, many algorithms and software tools have been developed to utilize the data contained in such event logs: By analyzing the execution of a process, as captured in the log, we get insights on deviations of the ideal process model, bottlenecks and waste of resources.

The field of process discovery focusses on extracting a model from a given event log, that reflects the process underlying the log: The observed events are put into relation to each other, preconditions, choices, concurrency, etc. are discovered, and brought together in a process model, e.g. a Petri net. Process discovery is non-trivial for a variety of reasons. Ideally, a discovered model should be able to produce the behavior contained within the event log (fitness), not allow for behavior that was not observed (precision), represent all relevant dependencies between the events and at the same time be simple enough to be understood by a human interpreter. It is rarely possible to fulfill all these requirements simultaneously. Based on the capabilities and focus of the used algorithm, the discovered models can vary greatly.

The recently developed process discovery algorithm eST-Miner and its variants return a process model in the form of a Petri net. This Petri net is constructed by considering the set of all possible places, then evaluating a smartly selected subset of these candidate places and finally inserting the good places into the returned model. The consideration of all possible places ensures that complex-control flow structures can be discovered. The individual evaluation of place candidates allows for an enormous flexibility when defining which places are to be selected for the final model. Heuristic evaluation of places based on a large variety of properties can help to speed up the algorithm by reducing the size of the candidate space, to quickly make decisions when places are conflicting, and to guide the discovered model to express particularly interesting aspects of the logged process.

In this thesis, the student implements a variety of place evaluation heuristics within the eST-Miner framework and performs an extensive experimental evaluation of performance and quality aspects using artificial and real-life event logs. A formal analysis of properties and guarantees including rigorous proofs may be required. The student presents the achievements in the context of their thesis paper as well as an intermediate and final presentation.

Prerequisites

This thesis includes implementation, testing and evaluation as well as proper presentation and discussion of the results, methods and background. Furthermore, the analysis of properties and guarantees may require formal reasoning and rigorous proofs. We expect good programming skills (Java), knowledge of basic computer science concepts, good theoretical foundations and formalization skills, and a strong interest in theoretical and practical aspects of process mining and Petri net theory.

Pointers

Supervisor

Prof.dr.ir. Wil van der Aalst

Advisor

Lisa Mannel

For more Information

Send an e-mail to . Make sure to include detailed information about your background and scores for completed courses.