TY - GEN
T1 - How much language is enough? Theoretical and practical use of the business process modeling notation
AU - Zur Muehlen, Michael
AU - Recker, Jan
PY - 2008
Y1 - 2008
N2 - The Business Process Modeling Notation (BPMN) is an increasingly important industry standard for the graphical representation of business processes. BPMN offers a wide range of modeling constructs, significantly more than other popular languages. However, not all of these constructs are equally important in practice as business analysts frequently use arbitrary subsets of BPMN. In this paper we investigate what these subsets are, and how they differ between academic, consulting, and general use of the language. We analyzed 120 BPMN diagrams using mathematical and statistical techniques. Our findings indicate that BPMN is used in groups of several, well-defined construct clusters, but less than 20% of its vocabulary is regularly used and some constructs did not occur in any of the models we analyzed. While the average model contains just 9 different BPMN constructs, models of this complexity have typically just 4-5 constructs in common, which means that only a small agreed subset of BPMN has emerged. Our findings have implications for the entire ecosystems of analysts and modelers in that they provide guidance on how to reduce language complexity, which should increase the ease and speed of process modeling.
AB - The Business Process Modeling Notation (BPMN) is an increasingly important industry standard for the graphical representation of business processes. BPMN offers a wide range of modeling constructs, significantly more than other popular languages. However, not all of these constructs are equally important in practice as business analysts frequently use arbitrary subsets of BPMN. In this paper we investigate what these subsets are, and how they differ between academic, consulting, and general use of the language. We analyzed 120 BPMN diagrams using mathematical and statistical techniques. Our findings indicate that BPMN is used in groups of several, well-defined construct clusters, but less than 20% of its vocabulary is regularly used and some constructs did not occur in any of the models we analyzed. While the average model contains just 9 different BPMN constructs, models of this complexity have typically just 4-5 constructs in common, which means that only a small agreed subset of BPMN has emerged. Our findings have implications for the entire ecosystems of analysts and modelers in that they provide guidance on how to reduce language complexity, which should increase the ease and speed of process modeling.
KW - BPMN
KW - Language analysis
KW - Process modeling
UR - http://www.scopus.com/inward/record.url?scp=45849132040&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=45849132040&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-69534-9_35
DO - 10.1007/978-3-540-69534-9_35
M3 - Conference contribution
AN - SCOPUS:45849132040
SN - 3540695338
SN - 9783540695332
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 465
EP - 479
BT - Advanced Information Systems Engineering - 20th International Conference, CAiSE 2008, Proceedings
T2 - 20th International Conference on Advanced Information Systems Engineering, CAiSE 2008
Y2 - 16 June 2008 through 20 June 2008
ER -