Generalization of Workflows in Process-Oriented Case-Based Reasoning

In this paper, we introduce the concept of generalized cases into process-oriented case-based reasoning. We present the formal foundations for the generalization of workflow cases as well as a new algorithm for generalizing semantic workflows, guided by ontological knowledge of the domain. Further, the specialization of workflows w.r.t. a current query is addressed. An experimental evaluation demonstrates the capability of the approach for workflow adaptation showing that the adapted workflows have a similar quality compared to that of original workflows. Furthermore, the retrieval performance can be improved by a reduction of the case-base size while the coverage of cases is significantly increased.