초록

AI-Helper

한국어의 구문 및 의미해석등에 필요한 정보를 도출하기 위한 입력분의 형태해석에 대하여 기술한다. 명사구에서 명사와 조사를 분리하고, 복합명사의 분리점을 선택하는 조건을 규정하며, 변형된 복합명사를 처리하는 규칙을 표시한다. 그리고 규칙동사에서 어간과 어미를 분리하고, 변칙용언과 음운축약 등을 효과적으로 처리하기 위한 논리표현 형식을 제안한다. 이 논리표현은 해석규칙과 속성값으로 구성한다. 명사사전의 중복을 배제하기 위하여 "명사형 하다" 동사의 분리 처리와 Q parameter 도입에 의한 "이다"의 처리방법을 보이고, 또한 부정문의 처리형식도 유도하여 복합술부의 형태소와 기본형식을 제시한다.

Abstract ▼ AI-Helper

A morphological analysis is described to extract the informations which are required in syntactic and semantic analysis of the Korean language. The noun and particle are separated in a noun phrase, the selecting conditions are specified to analyze the compound noun and a restoring rule is represente...

A morphological analysis is described to extract the informations which are required in syntactic and semantic analysis of the Korean language. The noun and particle are separated in a noun phrase, the selecting conditions are specified to analyze the compound noun and a restoring rule is represented to process the irregular compound noun. The stem and ending are separated in normal verbals and a logical representive form is proposed to the anomalously inflected word and contracted vowels. The logical representation is composed of the attribute value an analyzing rule. The redundancy of noun is reduced in the dictionary as the verb of a "Nounformed HA-" is processed by "noun" and "HA-", separately and a predicative "IDA" is analyzed by Q parameter. The processing form of negation is also derived and the morpheme and basic structure of compound predicative parts are presented.