Similarity Matching of Business Process Variants

Abstract

Evidence from business work practice indicates that variance from prescribed business process models is not only inevitable and frequent, but is in fact a valuable source of organizational intellectual capital that needs to be captured and capitalized, since variance is typically representative of preferred and successful work practice. In this paper, we present a framework for harnessing the value of business process variants. An essential aspect of this framework is the ability to search and retrieve variants. This functionality requires variants to be matched against a given criteria. The focus of this paper is on the structural criteria which is rather challenging as query process structures may have different levels of similarity with variant process structures. The paper provides methods for undertaking the similarity matching and subsequently providing ranked results in a systematic way, as well as a reference architecture within which the methods may be deployed.