Modification of Alpha++ for Discovering Collaboration Business Processes Containing Non-Free Choice

Abstract

A business process is a series of activities or work that is structured and interrelated to solve a problem. Business Process Model Notation (BPMN) is a standard that often use to create a series of business processes. There are not only one or two processes in a business process, but many processes which cross each other. Processes which have activities related to one another are called collaborative business processes. Process mining techniques analyze a process model that is generated based on actual processes with a process model as the Standard Operational Procedure (SOP). Existing process mining techniques depict activities of each process without consider the linkages between activities in a process with another process. In addition, there is an issue that should be handled by process mining techniques, i.e. non-fee choice. This paper proposes a Modification of Alpha++ for discovering linkages of activities in collaboration business processes. Alpha++ is chosen because this algorithm can form non-free choice relationship. In the evaluation, the proposed Modification of Alpha++ will be compared with Alpha++, Inductive Miner, and Heuristic Miner. The evaluation shows the proposed Modification Alpha++ algorithm can model event log data correctly based on the proposed business process. Modification Alpha++ can find non-free choice constructs on the event log data. Modification Alpha++ gets the best calculation results for fitness, precision, and f-measure with a value of 1.