초록

용어보기논문에서 용어와 풀이말을 자동 추출한 결과로, 시범 서비스 중입니다. AI-Helper

Abstract ▼ AI-Helper

This paper proposes a two-step extended model for Korean morphological analysis considering the efficiency of the word recognition algorithm, the acceptability of analysis result, and various requirement specifications for application area. Two-step extended model consists of word recognition step ...

This paper proposes a two-step extended model for Korean morphological analysis considering the efficiency of the word recognition algorithm, the acceptability of analysis result, and various requirement specifications for application area. Two-step extended model consists of word recognition step using basic dictionary and morpheme extension step using extended dictionary. In the word recognition step, lexical part and grammatical part are recognized according to the simplified parts of speech (POS) such as noun/verb/others and simplified word pattern like noun+Josa, verb+Eomi, and single word. In the word extension step, lexical part and grammatical part are extended by the subdivided word pattern and POS pattern. Furthermore, priority of the analysis result are given by the heuristic rules that are easily found in the morphological analysis procedure and application-dependent results are generated. Korean morphological analyzer implemented by this model can easily analyze a complicate word by top-down predictive analysis method. The requirements of automatic word spacing in the speller, unknown word guessing and decomposition of compound nouns for automatic indexing, and specific information about morpheme in the Korean understanding system are easily reflected in the morphological analyzer because morphological processing is divided into kernel and extended function.