Title :
An Empirical Analysis of Business Process Execution Language Usage
Author :
Hertis, Matej ; Juric, Matjaz B.
Author_Institution :
Lab. for Integration of Inf. Syst., Univ. of Ljubljana, Ljubljana, Slovenia
Abstract :
The current state of executable business process languages allows for and demands optimization of design practices and specifications. In this paper, we present the first empirical study that analyses Web Services Business Process Execution Language (WS-BPEL or BPEL) usage and characteristics of real world executable business processes. We have analysed 1,145 BPEL processes by measuring activity usage and process complexity. In addition, we investigated the occurrence of activity usage patterns. The results revealed that the usage frequency of BPEL activities varies and that some activities have a strong co-occurrence. BPEL activities often appear in activity patterns that are repeated in multiple processes. Furthermore, the current process complexity metrics have proved to be inadequate for measuring BPEL process complexity. The empirical results provide fundamental knowledge on how BPEL specification and process design practices can be improved. We propose BPEL design guidelines and BPEL language improvements for the design of more understandable and less complex processes. The results are of interest to business process language designers, business process tool developers, business process designers and developers, and software engineering researchers, and contribute to the general understanding of BPEL and service-oriented architecture.
Keywords :
Web Services Business Process Execution Language; service-oriented architecture; BPEL activities; BPEL design guidelines; BPEL language improvements; WS-BPEL; Web services business process execution language; activity usage; design practices; design specifications; empirical analysis; executable business processes; process complexity; service-oriented architecture; Business; Complexity theory; Guidelines; Measurement; Semantics; Syntactics; XML; WS-BPEL Analysis; complexity measure; empirical study; process complexity; process comprehension; process patterns; service composition;
Journal_Title :
Software Engineering, IEEE Transactions on
DOI :
10.1109/TSE.2014.2322618