비즈니스 프로세스 도출 알고리즘 추천 프레임워크 및 발견적 규칙 기반 프로세스 도출 알고리즘 개발

Abstract

Under pressure of the rapid change engendered by the fast growth of information and communication technologies, organizations need to continuously enhance their business processes in order to defend their market position and maintain their competitive edge. To achieve this, process mining has emerged as a mean to analyze the behavior of companies. Business process mining is new methods that amalgamate business process modeling and analysis with data mining, artificial intelligence and machine learning techniques, whereby process-oriented knowledge from event logs stored in today’s information systems are extracted to automatically discover business process models, identify bottlenecks, and improve the business processes. Many powerful process discovery algorithms have recently been developed. However, users and businesses still cannot choose or decide the appropriate mining algorithm for their business processes. Each algorithm has a specific limitation regarding the mining of short loops, invisible tasks, duplicate tasks and non-free choice constructs. There is no algorithm which is capable of discovering the aforementioned characteristics in a restricted time if all of them are present in the event log. The goal of this research consists of first developing a process discovery algorithms recommendation framework capable of recommending to businesses the most suitable process discovery technique to their business processes based on the knowledge in the event logs of the processes in question; second developing a new process discovery algorithm capable of handling standard constructs, short loops, invisible tasks, duplicate tasks, and non-free choice constructs if all of them exist in the event log.