A Process Mining-Based Solution for Business Process Model Extension with Cost Perspective Context-Based Cost Data Analysis and Case Study

Abstract

Several organizations look for improving their business processes in order to enhance their efficiency and competitiveness. The lack of integration between the business process model and its incurred financial cost information hampers for better decision making support allowing business process incurred cost reduction. In previous work, we proposed a solution for business process model extension with cost perspective based on process mining, independently of the business process model notation. The proposed solution provides cost data description and analysis at the process and the activity levels. Cost data analysis allows to extract knowledge about factors influencing on cost at each of the process and the activity levels. The proposed solution also involves cost data analysis through the use of classification algorithms which can be selected by the user. However, the lack of support during this selection may affect the accuracy of the obtained results. Furthermore, the performance of the same classification algorithm may vary from a case to another depending on its context: (1) data features and (2) the considered performance criteria. Thus, in this paper, we propose to adopt a context-based cost data analysis allowing to select and apply the classification algorithm the most suited to the case in hand. This supports improving the accuracy of the obtained results. In order to validate the proposed solution, a case study is conducted on the business process of a maternity department in a Tunisian clinic. The results of this case study confirm the expected goals.