Skip to main content

TS

Increasing recall of process model matching by improved activity label matching

Authors

Christopher Klinkmueller, Ingo Weber, Jan Mendling, Henrik Leopold and Andre Ludwig

University of Leipzig

NICTA

UNSW

Wirtschaftsuniversitaet Wien

Humboldt-Universitaet zu Berlin

Abstract

Comparing process models and matching similar activities has recently emerged as a research area of business process management. However, the problem is fundamentally hard when considering realistic process collections: there are numerous options of choosing terms of grammatical structure for activity labels, and different modelers often work on different levels of abstraction. While prior research has established important conceptual foundations, recall values have been fairly low (around 0.26) { arguably too low to be useful in practice. In this paper, we aim at improving recall by focusing on the essential first step of matching activity labels. Using various linguistic and semantic techniques, we achieve significant improvements (e.g., recall of 0.44, without sacrificing precision), as show-cased by a comparative evaluation. Furthermore, we identify categories of matching challenges to guide future research.

BibTeX Entry

  @inproceedings{Klinkmueller_WMLL_13,
    author           = {Klinkmueller, Christopher and Weber, Ingo and Mendling, Jan and Leopold, Henrik and Ludwig, Andre},
    year             = {2013},
    month            = aug,
    keywords         = {bpm, process similarity, process model matching},
    title            = {Increasing Recall of Process Model Matching by Improved Activity Label Matching},
    booktitle        = {International Conference on Business Process Management},
    pages            = {211-218},
    address          = {China}
  }

Download

Served by Apache on Linux on seL4.