15세기 문헌자료의 전산화 - 연구자 중심 말뭉치 구축 및 활용 -

Abstract

The current study aims to introduce how to computerize the original to utilize on various studies, choosing and sorting the materials by the wants of researchers in terms of vocabulary, word phrasing, and sentence structure required for their individual studies of 15th century Korean documents. In the case of computerizing 15th century Korean documents, it is certain that errorless corpus is the best corpus. However, as corpus construction work is time consuming, it is extremely difficult to create an errorless corpus. Therefore, when changing 15th century Korean documents into raw corpus, comparing and editing them with the original Korean documents should always come first. The ways to utilize this study with research processing corpus, within the topic or domain of what researchers want, are as follows. First, on Korean language materials, it is better to ensure that the time based meaning given by the important ‘book title’ and ‘publication period’ are processed together. Second, in the case of building a database for the diachronic inquiry of the change in Korean vocabulary, it must be based on processing with at least three pieces of information, that is to say, vocabulary information including the age information with the first book title to appear, similar vocabulary information processing with variations on the initial vocabulary, and information processing on the present meaning. In sum, if similar vocabulary(variation) of search vocabulary is extracted, from time information such as ‘book title’ and ‘publication period,’ and corpus listed by time and literature is built, it will have a value that can be used in various ways as not only with the 15th century Korean documents, but also with the research materials which Korean researchers want to use.