OBJECTIVES: To discuss the relationships between ontologies, terminologies and language in the context of Natural Language Processing (NLP) applications in order to show the negative consequences of confusing them. METHODS: The viewpoints of the terminologist and (computational) linguist are developed separately, and then compared, leading to the presentation of reconciliation among these points of view, with consideration of the role of the ontologist. RESULTS: In order to encourage appropriate usage of terminologies, guidelines are presented advocating the simultaneous publication of pragmatic vocabularies supported by terminological material based on adequate ontological analysis. CONCLUSIONS: Ontologies, terminologies and natural languages each have their own purpose. Ontologies support machine understanding, natural languages support human communication, and terminologies should form the bridge between them. Therefore, future terminology standards should be based on sound ontology and do justice to the diversities in natural languages. Moreover, they should support local vocabularies, in order to be easily adaptable to local needs and practices.
OBJECTIVES: To discuss the relationships between ontologies, terminologies and language in the context of Natural Language Processing (NLP) applications in order to show the negative consequences of confusing them. METHODS: The viewpoints of the terminologist and (computational) linguist are developed separately, and then compared, leading to the presentation of reconciliation among these points of view, with consideration of the role of the ontologist. RESULTS: In order to encourage appropriate usage of terminologies, guidelines are presented advocating the simultaneous publication of pragmatic vocabularies supported by terminological material based on adequate ontological analysis. CONCLUSIONS: Ontologies, terminologies and natural languages each have their own purpose. Ontologies support machine understanding, natural languages support human communication, and terminologies should form the bridge between them. Therefore, future terminology standards should be based on sound ontology and do justice to the diversities in natural languages. Moreover, they should support local vocabularies, in order to be easily adaptable to local needs and practices.