Demystifying ontological classification in language engineering

Abstract

© Springer International Publishing Switzerland 2016. The introduction of ontological classification to support domain-metamodeling has been pivotal in the emergence of multi-level modeling as a dynamic research area. However, existing expositions of ontological classification have only used a limited context to distinguish it from the historically more commonly used linguistic classification. In important areas such as domain-specific languages and classic language engineering the distinction can appear to become blurred and the role of ontological classification is obscured, if not fundamentally challenged. In this paper we therefore examine critical points of confusion regarding the distinction and provide an expanded explanation of the differences. We maintain that optimally utilizing ontological classification, even for tasks that traditionally have only been viewed as language engineering, is critical for mastering the challenges in complex systems modeling including the validation of multi-language models.

A Novel Approach to Modeling Enterprise Services Leveraging Object Cloning and Multilevel Classification

“…being either at the class level or the object level, and therefore suffer from what advocates of multilevel modeling call shallow instantiation [4]. Due to this inadequate granularity of representation, traditional two-level classification techniques does not work with deep instantiation typically found in domain ontologies and linguistic structures [21]. To remedy this, conceptual modeling is enhanced with the notion of clabject that is supposed to be instantiated for multiple times until its instantiation potency runs out [22].…”