Framework for Process Mining in Semi-structured Information Systems

Abstract

There are many application systems in the developing world that do not adhere to the systematic software development procedures and are not process-aware. Many of these systems do not meet their objective of efficiency & effectiveness. However, these applications can be analyzed and improved using information derived from event logs. We discuss what process mining is, how to extract & prepare event logs from systems and the different tools that are used to process the data and give meaningful information, how to analyse the results and eventually introduce a framework that can be used to go through process mining. We begin by following a framework developed by R.S. Mans, M.H. Schonenberg, M. Song, W.M.P. van der Aalst, and P.J.M. Bakker. However, the special circumstances of semi-structured systems are looked into, leading to changes in the framework. The proposed framework considers three main analysis methods to be performed on the event logs; control flow analysis, performance analysis, and role analysis. Data collection and preparation are given prominence due to the nature of discovery that is done on semi-structured systems. These systems in many cases will not have had previous statistical information of this crucial information. The process mining plugins help in getting this information. This research can be used to assist many organizations in cleaning up their systems and coming up with optimal processes which meet their objectives.