Method for mining hybrid multi-concurrent structure based on Petri network to improve efficiency of business process

Abstract

The invention belongs to the technical field of process mining, particularly discloses a method for mining a hybrid multi-concurrent structure based on a Petri network to improve business process efficiency, and aims to solve the technical problem of low business process efficiency due to the fact that an existing mining method cannot accurately mine a process model of the hybrid multi-concurrentstructure from incomplete event logs. The process mining method based on the Petri network, namely the AlphaMining method, comprises the following steps that firstly, all activities in an incomplete event log are obtained in an event system, and the activity conforming to the simplest second-degree short cycle structure is found out; respectively redefining the activities in the simplest two-degree cycle structure; secondly, according to the number characteristics and the position characteristics of the activities in the traces, correctly matching and identifying the same type of second-degreeshort-cycle structure by using a number marking matrix and a position marking matrix, and completing mining of a process model containing the hybrid multi-concurrent structure; and then, the mined process model is used for replaying the actual business process, so that the efficiency of the actual business process is improved.

Description

technical field [0001] The invention relates to a method for improving the efficiency of business processes by mining a mixed multi-concurrent structure based on Petri nets. Background technique [0002] Process mining is an emerging discipline that has provided a full set of tools to gain insight into facts and support process improvement in the past ten years. It is developed on the basis of process model-driven methods and data mining. The rapid development and application of process mining technology benefits from the increasing ease of obtaining event data from modern information systems. [0003] The core of process mining is to discover, comply with and improve the actual process by extracting knowledge from the event logs generated by the actual process. The discovery technique is to use the event log generation model that does not include any prior information, the compliance detection is used to check the deviation between the actual execution recorded in the log ...

AI Technical Summary

Problems solved by technology

[0014] By analyzing the current common process mining methods, incomplete logs that do not contain explicit cycle features similar to "aba" cannot effectively mine mixed multi-concurrent short cycle structures, making the obtained process model fit , accuracy and succinctness are poor, unable to meet the needs of the actual business process update, thus reducing the efficiency of the actual business process

Images