Complexity-Aware Generation of Workflows by Process-Oriented Case-Based Reasoning

Abstract

One of the biggest challenges in business process management is the creation of appropriate and efficient workflows. This asks for intelligent, knowledge-based systems that assist domain experts in this endeavor. In this paper we investigate workflow creation by applying Process-Oriented Case-Based Reasoning (POCBR). We introduce POCBR and describe how it can be applied to the experience-based generation of workflows by retrieval and adaptation of available best-practice workflow models. While existing approaches have already demonstrated their feasibility in principle, the generated workflows are not optimized with respect to complexity requirements. However, there is a high interest in workflows with a low complexity, e.g., to ensure the appropriate enactment as well as the understandability of the workflow. The main contribution of this paper is thus a novel approach to consider the workflow complexity during the workflow generation. Therefore, a complexity measure for workflows is proposed and integrated into the retrieval and adaptation process. An experimental evaluation with real cooking recipes clearly demonstrates the benefits of the described approach.