Ontology: buzzword or paradigm shift in GI science?

It is apparent that the term ontology is used recently more and more in the context of GI Science (Frank 1997, Smith and Varzi 1997, Kemp and Vckovski 1998, Smith and Mark 1998, Bittner and Winter 1999, Rodriguez et al. 1999, Fonseca et al. 2000, Timpf 2001). Is ontology then a buzzword, or a real paradigm shift? What is in it? Can ontology, for example, be the key to interoperability (Vckovski et al. 1999 )? Does it help to solve the open problems in data access and interpretation (semantic interoperability) , in mappings of meanings between application domains, in metadata modeling, in data quality issues, and in data mining? Does it improve our understanding of geographical space? Or is ontology a modern  owery phrase for former concepts, like formal data models, formal speciŽ cations, or semantic networks? Well, it depends. It depends mostly on the community using the term, which makes an interdisciplinary discussion di cult. For that reason Guarino distinguishes between Ontology in philosophy and ontologies in knowledge engineering (Guarino 1998). In the philosophical tradition, ontology is related to what exists a priori to perception, knowledge, or language. It is sometimes further divided in reality-based and epistemological ontology; the latter describes human conceptualizations of reality (Smith 1999). In knowledge engineering formal ontology copes with ( language dependent ) knowledge: it is used for ‘an explicit speciŽ cation of a conceptualization’ (Gruber 1993) or a ‘shared understanding of some domain of interest’ (Uschold and Gruninger 1996). Being confronted with the variety of uses of the term ontology I will not add another deŽ nition. Instead I Ž nd it useful studying what is going on under the label of ontology. The reader will Ž nd that even the papers collected in this issue diŒer in their understanding of ontology. It is up to the authors to deŽ ne their approach exactly. A synopsis turns out to be stimulating and promising: the work done in ontology and epistemology research is beyond conceptual modeling. The work asks for referential transparency or grounding of models; it makes the commitments to the world beyond the knowledge base itself explicit (Guarino and Poli 1995). It seems to be