This paper presents a translation approach to portable ontology specifications, introduced by Thomas R. Gruber. Ontologies are formal specifications of conceptualizations, which define the vocabulary for shared knowledge. The paper describes Ontolingua, a system that translates ontologies from a standard, system-independent format into specialized representation systems. This allows researchers to share and reuse ontologies while retaining the computational benefits of specialized implementations. Ontolingua is based on an ontology of domain-independent, representational idioms, which enables it to translate definitions written in a standard format for predicate calculus into specialized representations, including frame-based systems and relational languages. The paper discusses how the translation approach addresses technical problems such as accommodating stylistic and organizational differences among representations while preserving declarative content, and translating from a very expressive language into restricted languages while maintaining system independence and computational efficiency. The paper also discusses the role of ontologies in knowledge sharing among AI systems, emphasizing the need for common ontologies to support interoperability. It describes how ontologies can be used to define content-specific agreements for knowledge-sharing activities. The paper presents an example of an ontology for bibliographic information, showing how it defines classes, relations, functions, and axioms. It also discusses the translation of this ontology into different representation systems, such as Epikit, Loom, and KEE. The paper concludes by discussing the design of Ontolingua, which balances the need for an expressive, declarative, system-independent language with the need to support translation into specialized representations. It also discusses the Frame Ontology, which defines common knowledge-organization conventions used in object-centered or frame-based representations. The paper emphasizes the importance of representation ontologies in providing a framework for defining hierarchies of classes with slots, and the need for content ontologies to make claims about how the world should be described. The paper also discusses the limitations of Ontolingua, including its incompleteness with respect to the KIF language and the need for a common-denominator approach to ensure portability across specialized systems.This paper presents a translation approach to portable ontology specifications, introduced by Thomas R. Gruber. Ontologies are formal specifications of conceptualizations, which define the vocabulary for shared knowledge. The paper describes Ontolingua, a system that translates ontologies from a standard, system-independent format into specialized representation systems. This allows researchers to share and reuse ontologies while retaining the computational benefits of specialized implementations. Ontolingua is based on an ontology of domain-independent, representational idioms, which enables it to translate definitions written in a standard format for predicate calculus into specialized representations, including frame-based systems and relational languages. The paper discusses how the translation approach addresses technical problems such as accommodating stylistic and organizational differences among representations while preserving declarative content, and translating from a very expressive language into restricted languages while maintaining system independence and computational efficiency. The paper also discusses the role of ontologies in knowledge sharing among AI systems, emphasizing the need for common ontologies to support interoperability. It describes how ontologies can be used to define content-specific agreements for knowledge-sharing activities. The paper presents an example of an ontology for bibliographic information, showing how it defines classes, relations, functions, and axioms. It also discusses the translation of this ontology into different representation systems, such as Epikit, Loom, and KEE. The paper concludes by discussing the design of Ontolingua, which balances the need for an expressive, declarative, system-independent language with the need to support translation into specialized representations. It also discusses the Frame Ontology, which defines common knowledge-organization conventions used in object-centered or frame-based representations. The paper emphasizes the importance of representation ontologies in providing a framework for defining hierarchies of classes with slots, and the need for content ontologies to make claims about how the world should be described. The paper also discusses the limitations of Ontolingua, including its incompleteness with respect to the KIF language and the need for a common-denominator approach to ensure portability across specialized systems.