Development Of Ki Process Mining Plug-in And Alorithms Analysis For Patterns Extraction

In recent years, process analysis techniques are increasingly focused on the analysis of knowledge-intensive(KI) processes, which means processes whose value “can only be created by the implementation of the requirements knowledge regarding the participants in the process”.Domains such as emergency management,health care, or research as well as innovation development, are characterized by the performance of KI processes. They can be viewed as sequences of activities in accordance with knowledge intensive acquisition and treatment. These knowledge intensive processes can be executed in enterprises of different kind regardless of which type, production or service company.This paper proposed a behavior pattern extraction method, the purpose is to extract the relevant sub-processes.Firstly,a relatively formal definition of behavioral patterns is proposed. Secondly, to record the KI staff activities in the process, and use the log building method to form a event log.Next, research the heuristic clustering algorithm, and develop its’ plug-in to realize process mining to generates heuristic net, through plug-ins can achieve changing logs into data, extracting gained data to pattern extraction.Then developed PetriNet Replayer plug-in implements the conversion of mining results to output Petri nets. The results obtained for each plug-in can call visualization tool for graphical display. And gives the algorithm comparative analysis involves in the thesis. Finally, it has integrated developed plug-ins into one of the most prominent process mining research tools Pro M6, making Pro M platform more perfect.