Process mining method based on a mixed event log

Abstract

The invention discloses a process mining method based on a mixed event log, which takes a mixed event log including a point event and an interval event as an input, and takes a mining correct processmodel as an output. First, mining low-level event relationships from each event sequence in the log; then, based on all the low-level event relations, the event relations on the whole log level are deduced; finally, the model generation algorithm based on log-level event relation in the alpha algorithm is called to obtain the Petri net procedure. The invention can effectively mine the correct process model from the locally complete mixed event log, and compared with the existing mining method, is applicable to a wide range and is generally applicable to point event logs, interval event logs and mixed event logs.

Description

technical field [0001] The invention belongs to the field of data mining, and in particular relates to a process mining method based on a hybrid event log. Background technique [0002] With the continuous development of the Internet and big data applications, business process models play an increasingly important role in today's information systems, from traditional automated office systems, organizational business management systems to service portfolios and big data in cloud environments Applications, etc., the business logic of many applications can be abstracted into processes. At the same time, the business process has become unprecedentedly complex, so how to dig out high-quality process models has become an important challenge. [0003] Event logs, as an important part of computer information systems, record events that occur during system execution, and related information such as the time, operator, and life cycle of the events, which are crucial to understanding ...

AI Technical Summary

Problems solved by technology

[0005] Most of the existing process mining technologies are based on event logs that only contain point events or only interval events. However, most event logs in reality are of mixed type. The application of existing process mining technologies to mixed event logs requires log mining. Preprocessing, expanding the point event to an interval event, or deleting the start event or end event of the interval event, cannot guarantee the correct process model to be mined

Images