Engineering Ontologies vs . Linguistic Ontologies for Web Applications

semanticscholar(2001)

引用 0|浏览0
暂无评分
摘要
A classification of ontologies (Guarino 97) for information processing distinguishes the four different kinds of ontologies: • Top-level ontologies describe generic concepts independently of a particular domain or problem. A top-level ontology defines fairly abstract categories of time, space, event, action, etc. Typically, linguistic ontologies that encode a common sense view of the world fall under this category, for example Mikrokosmos (Mahesh 96) or Sensus (Knight & Luk 4). • Domain ontologies describe the terminology of a generic domain. A typical example is the UMLS Metathesaurus defined by the U.S. National Library of Medicine for the medical domain (National Library of Medicine 97). • Finally, an application ontology is a combination or specialization of task and domain ontologies, where entities may be characterized by their roles. Ontologies describe possible worlds. A knowledge base describes an instance of a possible world as a set of facts. The knowledge base repository records specific events, places, people, objects, etc., as classified by the ontology. A knowledge base can be structured (as a database for example), semi-structured (as an HTML or XML repository for example) or unstructured (a set of free texts). Although these distinctions are useful, they apply mostly to engineering ontologies, and we present in the following sections some problems that eschew this simple categorization. Engineering ontologies are application ontologies which are typically used at development time to This paper outlines the basic differences between language ontologies, engineering ontologies and Web ontologies such as Yahoo!'s directory. Each type of application has different requirements that are not always easily reconciled. Problems in ontology integration and standardization reinforce the view that the idea of a universal reusable ontology is perhaps illusory. The paper concludes by pointing at some directions which have been explored in the context of syntax and lexical semantics, but not applied to other domains.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要