摘要
本文论述了传统叙词表的语义关系及其所具有的局限性,对FrameNet叙词表的语义关系如词汇聚合关系、继承关系、总分关系、参照关系、因果与状态关系进行了充分揭示,并总结了其明显的语义特征。FrameNet叙词表运用本体论思想,将人工语言与自然语言、词汇与句法有机结合,为今后自动语义标注、问答系统、信息抽取和信息检索的应用奠定了基础。
With a discussion on the semantic relationships of traditional thesaurus and its limitations, this paper describes the semantic relationships of FrameNet thesaurus in depth such as lexical aggregation, inheritance, subframe, reference, and causality and state, and summarizes its distinct semantic characters. FrameNet thesaurus uses the theory of ontology and combines manual language with natural language, lexics with syntax, forming a basis for automatic semantic labeling, QA systems, information extraction and information retrieval.
出处
《情报理论与实践》
CSSCI
北大核心
2006年第5期605-607,576,共4页
Information Studies:Theory & Application
