Merging event logs: Combining granularity levels for process flow analysis

Lihi Raichelson, Pnina Soffer, Eric Verbeek

Research output: Contribution to journalArticlepeer-review

Abstract

Process mining techniques enable the discovery and analysis of business processes and the identification of opportunities for improvement. Processes often comprise separately managed procedures documented in separate log files, which are impossible to mine in an integrative manner as the complete end-to-end process flow is obscure. These procedures can have simple (one-to-one) or complex (many-to-one or many-to-many) relationships among them. When complex relationships exist, typically different granularity levels are involved. In this paper, we present a merging algorithm that results in a comprehensive merged log that can handle all kinds of relationships between the procedures. Addressing differences in the granularity levels, it offers two views of the end-to-end process: a case view and an instance view. This enables the identification of process flow problems that could not be detected by previous techniques. The unified log can be used by process mining techniques to identify flow problems, particularly at the point of integration between the processes under consideration. The procedure proposed in this paper has been implemented and evaluated using both synthetic and real-life logs.

Original languageEnglish
Pages (from-to)211-227
Number of pages17
JournalInformation Systems
Volume71
DOIs
StatePublished - Nov 2017

Bibliographical note

Funding Information:
The research was partly supported by the Israel Science Foundation under grant agreement No. 856/13 .

Publisher Copyright:
© 2017 Elsevier Ltd

Keywords

  • Abstraction level
  • End-to-end process flow
  • Merging logs
  • Multiple instances
  • Process mining

ASJC Scopus subject areas

  • Software
  • Information Systems
  • Hardware and Architecture

Fingerprint

Dive into the research topics of 'Merging event logs: Combining granularity levels for process flow analysis'. Together they form a unique fingerprint.

Cite this