자연언어처리 기반 한·중 대응 문법구조의 중요성 고찰 ― ‘있다’를 중심으로

Abstract

With advancements in information and communication technology, there is a growing importance of natural language processing (NLP) in the development of language processing systems, such as machine translation, speech recognition, and linguistic database construction. Thus, this study explored methods to increase the accuracy of the NLP of Korean-Chinese corresponding grammatical structures, focused on the Korean word “있다”. The Korean-Chinese corresponding structures were analyzed with focus on the Korean word “있다” and its corresponding Chinese characters: “有,” “在” and “着”. This study translated related sentences using Google Translate, analyzed error types and organized grammatical structures. After that, erroneous sentences related to “있다” were translated using Google Translate, and the translated sentences were analyzed. Based on the findings, this study suggested methods to correct and adjust errors. The improvement methods drawn from the analysis of translated sentences can be summarized as follows. Improvement method 1: Provide a rule about spacing words in Korean Improvement method 2: Provide a rule about corresponding words e.g.) 서울=首尔，오빠=哥哥 Improvement method 3: Provide a rule about Korean-Chinese corresponding grammatical structures e.g.) Chinese: numeral + classifier + noun, Korean: noun + numeral + classifier It is hoped that this study will help establish a computerized system for Korean and Chinese.