Article ID Journal Published Year Pages File Type
4965593 Computers in Industry 2017 16 Pages PDF
Abstract
Industrial enterprises embody a rather large and heterogeneous business process landscape including hundreds to thousands of both manufacturing and supporting processes. To keep records of their business process architecture enterprises use informal and as well formal process descriptions. Formal process descriptions are often referred to as process models and are gaining increasing importance as a basis for process-aware enterprise information systems and automation purposes. Therefore process models are continuously adapted to changing business requirements. Keeping track of model changes is an important requirement to be able to understand past decisions and their impact on the process landscape. Hence, keeping track of changes is not easy if changes are not associated with the original rationale and the order of atomic changes is not preserved anymore. In this paper we present an approach that builds upon the concept of change patterns. For this purpose we systematically examined revision histories from a large process model collection and described them through a pattern language. In addition, we propose an algorithm to detect such change patterns. Our approach has been implemented in a modeling environment and has been evaluated with regard to effectiveness and performance.
Related Topics
Physical Sciences and Engineering Computer Science Computer Science Applications
Authors
,