A Translation Approach to Portable Ontology Specifications

To support the sharing and reuse of formally re presented knowledge among AI systems, it is useful to define the common vocabulary in which shared knowledge is represented. A specificat ion of a represent a t i onal vocabulary fo r a shared domain of discourse — definitions of classes, relations, functions, and other objects — is called an ont o logy. This paper describes a m e chanism for defining ont o logies t h at are p o rt ab le over representation systems. Definitions writ t e n in a st andard form at for p r edicat e calcul us are t r anslat ed b y a syst em called Ont o lingua int o specialized represent a t i ons, including frame- based systems as well as relational languages. This allows researchers t o share and reuse ont o lo gies, while ret a ining t h e comput at ional benefit s of specialized implementations. We discuss how t h e t r anslat ion approach t o po rt abilit y addresses several t e chnical problems. One problem is how t o accommodat e t h e st ylist i c and organizat ional differences among representations while preserving declarative conten t. Another i s how to tra n sl a t e f r om a very expressive language into restricted languages, remaining system-independent while preserving the computational efficiency of implemented sy stems. We describe how these problems are addressed by basing Ontolingua itself on an on tology of domain-independent, representational idioms.