Process mining methode voor het analyseren van performance business rules van het sollicitatieproces bij een uitzendbureau

Abstract

rocess mining uses event logs to analyse how a process is running, which can be used to better understand the performance of the process. Business rules are a logical expressions of process constraints within the process model. In this research, an approach was designed for the use of process mining and business rules to analyse the performance of the application process of an employment agency.

The performance business rules were not available; therefore, a process mining approach with a preparation phase was sought. We chose to extend the PM2 methodology with semi-structured interviews to retrieve business rules applicable to process mining. The interview questions were designed around the process mining filters of the process mining tool Disco. The designed interviews successfully retrieved 12 business rules from process experts and also provided input for creating a process model. This is useful because the process model can show where the constrains of the business rules affect the process.

In the process mining phase, the designed business rules were used as filters in Disco to divide process runs (cases) into cases that conform to the desired performance and cases that do not conform to the desired performance. A total of 10 business rules were applicable for process mining and provided an indication of the current process performance at the business unit and agency level, in accordance with the aim of the approach.

This research provides an evidence that the use of process mining to analyse the performance of a business process demands interviews or other research forms to retrieve business rules of the studied business process. The set of interview questions proposed in this research can be reused for retrieval of business rules of other business processes.

Date of Award	17 Feb 2021
Original language	Dutch
Supervisor	Ella Roubtsova (Examiner) & Lloyd Rutledge (Co-assessor)