Knowledge Acquisition from EPC Models for Extraction of Process Patterns in Engineering Domains

This paper presents an approach for the automated extraction of process patterns from Event-driven Process Chain (EPC) models in engineering domains. The manually extraction of process patterns (semantically described reference building blocks) is a labor-intensive, tedious and cumbersome task. The introduced approach comprises the three stages knowledge acquisition, process pattern extraction and generic pattern construction that are conducted automated. The presented approach is characterized by exploiting the implicit semantics of natural language terms associated with EPC model elements. Semantic patterns are employed to analyze the lexical structure of these terms and to relate them to instances of a reference ontology. Thus semantically annotated EPC model elements are input for subsequent process pattern extraction. The semantic annotation allows during pattern extraction to identify process goals of EPC functions and to organize goals into hierarchies. During generic process pattern construction, common goals of different process patterns give rise to construct generic process patterns.