SePMa: An algorithm that mining sequential processes from hybrid log

Abstract

To accommodate ourselves to the changeful and complex business environment, we should be able adjust the business processes within the enterprise whenever changes happen. However, the work to design and redesign the processes is far from trivial, the designers are required to have deep knowledge of the business processes at hand, in traditional approaches it means long term investigation and high cost. To automate the procedure of process discovery, process mining is introduced. Process mining takes the run-time log generated by the process management system as its input, and outputs the process models defined for the system. Unfortunately, current work on process mining often assumes that the input log is generated by the same process, but in many occasions this requisition is hard to be satisfied. In this paper, we propose SePMa, an algorithm that mining sequential processes from hybrid log. SePMa aims at discovering sequential processes from log generated by multiple processes, both of theoretical analysis and experimental results show that SePMa has very high efficiency and effectiveness. © Springer-Verlag Berlin Heidelberg 2007.