N+N 복합 구성의 자료 구축 및 분석 연구-모두의 말뭉치와 한국어 교재의 코퍼스를 바탕으로-

Abstract

This paper is a basic study for selecting ‘N+N’ composite structures for Korean language education, and aims to build a list of composite structures extracted from actual corpora (모두’s Corpus, Korean Language Textbook Corpus). To this end, we analyzed the frequency rankings of ‘모두’s Corpus’, which is actual language data, and sub-corpora, and analyzed the frequency rankings and overlap rankings between textbooks of Korean language textbook corpora that reflect the Korean language education field. In addition, we compared and analyzed the frequency ranking lists between the two corpora (including sub-corpora). The research results are as follows. First, we secured and constructed a list of frequency rankings of individual corpora (모두's Corpus, Textbook Corpus). Second, we compared the lists between individual corpora (spoken corpus and written corpus) and found that there was a significant difference in the high-frequency lists of the two corpora. We also confirmed the difference in the lists between the '모두's Corpus', a general corpus, and the 'Textbook Corpus', which reflects the Korean language education field. Third, we calculated the overlap between corpora, which is an indicator other than frequency, and confirmed the difference between corpora. Through this study, we confirmed that in order to secure meaningful lists in the future, we should pay attention not only to the overall frequency ranking but also to the frequency ranking of individual corpora, and that we should pay attention not only to high-frequency lists but also to lists that are used with high duplication in various genres.