Mereotopology: A Theory of Parts and Boundaries

The term ‘ontology’ has recently acquired a certain currency within the knowledge engineering community, especially in relation to the ARPA knowledge-sharing initiative (see Gruber (to appear), Mars (ed.) 1994, Guarino 1994, Guarino, Carrara and Giaretta 1994, 1994a). The term is used in a number of different senses, however, not all of them clear or mutually compatible. Here I follow philosophical tradition in conceiving ontology as the science which deals with the nature and the organisation of reality. Ontology thus conceived may be formal, in the sense that it is directed towards formal structures and relations in reality. This formal ontology is contrasted with the various material ontologies (of physics, chemistry, medicine, and so on) which study the nature and organisation of specific sub-regions of reality. Formal structures, for example the structures governing the relation of part to whole, are shared in common by all material domains. Both formal and material ontologies may be pursued with the aid of the machinery of axiomatic theories, and it is axiomatic formal ontology that has proved to be of most interest for the ontology-building purposes of the knowledge engineer.