Process Mining Possibilities and Challenges: A Case Study

Abstract

With today’s highly automated business environment and exponential growth of diverse data, new research domains are being established. An example is process mining, a discipline that emerged in the last decade and has process-centric view on data science. Process mining uses real event data in the form of event logs, generated usually from extracted event data of Process-Aware Information Systems (PAIS), in order to automatically construct business process models, compare them with an event log of the same process and improve existing process models, therefore, bridging the gap between process model analysis and data-oriented analysis. One of the main challenges of process mining, identified by the IEEE Task Force on Process Mining, is the lack of representative benchmarks and the consensus on process mining project methodology. In this paper, based on the comparison of two most prominent process mining tools, against their technical and performance features, a solution for conducting discovery and enhancement types of process mining, alongside with social network mining is presented. Furthermore, provided case study applies proposed solution on publicly available manufacturing event log extracted from ERP system, performed in both ProM and Disco process mining tools and presents the discovered process model, social network and detected problems.