Abstract

In this paper, we address the classification dependency used to illustrate that one concept is an instance-of another concept. In doing so, we present four important aspects of the classification dependency in the integration of implementation-neutral Karlstad Enterprise Modelling schemata. First and second, the classification dependency can be used not only to recognise and resolve power types but also to recognise and resolve homonyms. Third, in inference rules, the classification dependency can be used to deduce both new concepts and dependencies from already existing ones. Fourth, the classification dependency can be used to counter the occurrence of semantic loss and maintain the vocabulary used in the source schemata. The classification dependency can even be used to semantically enrich the integrated conceptual schema. The four aspects of the classification dependency should also be viewed as important aspects to take into account during the development of a semi-automatic method for the integration of implementation-neutral structural EM schemata. Finally, by applying the classification dependency, several problems, such as power types and homonyms, might be recognised that otherwise could pass unnoticed in the integration process.

Share

COinS