Designing a Process Mining-Enabled Decision Support System for Business Process Standardization in ERP Implementation Projects

摘要

Process standardization allows to optimize ERP systems and is a necessary step prior to ERP implementation projects. Traditional approaches to standardizing business processes are based on manually created "de-jure" process models, which are distorted, error-prone, simplistic, and often deviating from process reality. Theoretically embedded in the organizational contingency theory as kernel theory, this paper employs a design science approach to design a process mining-enabled decision support system (DSS) which combines bottom-up process mining models with manually added top-down standardization information to recommend a suitable standard process specification from a repository. Extended process models of the as-is process are matched against a repository of best-practice standard process model using an attribute-based process similarity matching algorithm. Thus, the DSS aims to reduce the overall costs of process standardization, to optimize the degree of fit between the organization and the implemented processes, and to minimize the degree of organizational change required in standardization and ERP implementation projects. This paper implements a working prototype instantiation in the open-source process analytics platform Apromore based on a real-life event log and standardization attributes for the Purchase-to-Pay and Order-to-Cash processes from three SAP R/3 ERP systems at the industry partner.