Mining batch activation rules from event logs

Niels Martin, Andreas Solti, Jan Mendling, Benoit Depaire, An Caris

Publication: Scientific journalJournal articlepeer-review

Abstract

Batch processing refers to an organization of work in which cases are synchronized such that they can be processed as a group. Prior research has studied batch processing mainly from a deductive angle, trying to identify optimal rules for composing batches. As a consequence, we lack methodological support to investigate according to which rules human resources build batches in work settings where batching rules are not strictly enforced. In this paper, we address this research gap by developing a technique to inductively mine batch activation rules from process execution data. The obtained batch activation rules can be used for various purposes, including to explicate the real-life batching behavior of human resources; to determine the compliance between the prescribed and actual batching behavior; or to investigate the influence of alternative batching behavior on service levels. The evaluation of our technique using both synthetic and real-world data demonstrates its effectiveness. With this work we complement prescriptive research on batch processing with a descriptive technique that is empirically grounded in process execution data.
Original languageEnglish
Pages (from-to)1
JournalIEEE Transactions on Services Computing
DOIs
Publication statusPublished - 2019

Cite this