Christopher van Dun won CDQ Young Talent Award
Christopher van Dun won the CDQ Young Talent Award for his excellent Master thesis!
The CDQ Young Talent Award recognizes an outstanding Master or Bachelor thesis in corporate data management. Christopher’s thesis – “Quality-Informed Semi-Automated Event Log Generation” – was written at the University of Augsburg in collaboration with the Queensland University of Technology. He convinced the top-class jury by translating data quality concepts to the domain of process mining. Following design science research, he implemented an approach to operationalize data quality dimensions and integrate data quality checks in the preparation of event logs.
Here you can find a short summary of his thesis:
The thesis examines a phase of the typical cross-industry standard process for data mining (CRISP-DM), the importance of which is easily underestimated: data preparation. For process mining, a subdomain of the more general field of data mining, input data needs to be prepared in a flat table structure (called event log). However, typical business information systems store data in relational databases and are not process-aware, i.e. they do not explicitly store process data. Instead, process data (including events and their respective timestamps) needs to be identified and extracted from relational tables. This work is often performed manually and therefore prone to errors. However, the quality of process mining results, as in most data mining applications, depends heavily on the quality of the input data.
Therefore, Christopher’s thesis presents an approach called RDB2Log, which offers support in analyzing the quality and suitability of the relational data for process mining. By analyzing the quality of the source data, the approach can provide meaningful insights into the suitability of a given data attribute for a specific role in a target event log. Christopher has implemented the approach as a software prototype that takes as input a relational database and offers users a guided process through which they can build event logs that suit their needs and are of high quality at the same time. The theoretical approach as well as the prototype instantiation have been evaluated thoroughly with experts in process mining and data quality.
At the awards ceremony during a workshop of the Competence Center Corporate Data Quality (CC CDQ) at Beiersdorf AG's corporate headquarters in Hamburg, Christopher presented a summary of his work. The audience consisted of data management experts from CC CDQ member companies such as Bosch, Merck, Schaeffler, and SAP who were very impressed by the innovative contribution. Christopher is currently developing his concept further and is preparing a scientific publication. We wish him a successful career!