Local log sampling method for process discovery

Abstract

The invention discloses a local log sampling method for process discovery. According to the method, the direct following relation between the events and the feature attributes of the trajectory are quantified, whether the trajectory has a new process behavior or not is taken as a sampling standard, and the minimum number of continuous traversal samples is determined based on a statistical theory. In order to further improve the preprocessing speed, a binary index jump algorithm is provided to avoid scanning of repeated tracks. By adopting the log sampling method disclosed by the invention, the scale of the event log can be quickly and effectively reduced, key control flow and frequency information are reserved, and meanwhile, the running speed of a process discovery algorithm is improved.

Description

technical field [0001] The invention belongs to the field of business process management, and aims at the preprocessing problem of large-scale event logs, and proposes a local log sampling method that can be used for process discovery. Background technique [0002] Process mining is a discipline that aims to extract business process-related insights from the information management system of an enterprise organization. Its core concept is to discover, monitor and improve real business processes. Process discovery is a subfield of process mining, and its main task is to build process models from event logs to achieve better understanding and analysis of business processes. [0003] Event logs are the basis for extracting process models. In recent years, with the rapid development of technologies such as the Internet of Things and big data, enterprises have more means to obtain and store event logs, resulting in a geometric increase in the size of event logs. Existing process...

AI Technical Summary

Problems solved by technology

However, these process mining algorithms often have multiple adjustable parameters, making process discovery a lengthy exploratory process, especially when applying them to large-scale event logs, which may require long processing time, resulting in Process discovery inefficiencies

Images