An Experimental Analysis on Mining Proportional Process Models from Process Logs

Joo-Eun Bae, Jin-Hyong Lee, Young-In Park, So-Hyang Park, Kyoung-Sook Kim, Kwanghoon Pio Kim

Abstract: In this paper, we carry out an experimental analysis to show how much perfectly the conceptual mining framework is working on mining process patterns and their enacted proportions from the business process enactment event histories logged with the IEEE-XES standardized format. In principle, the framework must be able to properly handle all the process patterns based upon the four types of control-flow primitives, such as linear (sequential), disjunctive (selective), conjunctive (parallel), and iterative (loop) patterns, together with the execution proportions. The paper focuses on not only verifying the conceptual feasibility of the procedural process mining framework through a series of experimental activities by using the implemented algorithmic mining framework and system, but also proving the functional correctness of the implemented process mining framework by applying to the real process instance enactment event histories of 10,000 workcases instantiated from the Large Bank Transaction Process Model and visualizing the details of the analytical artifacts and results, as well.