만주어 코퍼스의 구축 — ≪滿文老檔≫ 태조편을 대상으로 —

Abstract

By using Manwen laodang Taidzu as the target material, this study seeks a way to construct a morphologically annotated corpus for linguistic and cultural research. The Manchu corpus constructed in this study is an annotated corpus, and the annotation level is set to morphological annotation. In this study, we construct a morphologically annotated corpus by building a rule-based stemmer and tagging system using existing dictionaries, collectible vocabulary lists, and language knowledge of Manchu experts. Manwen laodang electronic document for annotated Manwen laodang corpus construction is Kim et al. (2019) was used. The purpose of the system constructed in this study is to produce a corpus that can be used for research by accurately tagging a large amount of data, which is a historical corpus. In other words, the goal of this study is to produce an annotated corpus by applying various heuristics to build a reusable annotated corpus. We anticipate that different approaches will also be used in Manchu and Altaic studies. Furthermore, we anticipate that the corpus released as a result of this study will serve as the basis for data centered Manchu language research.