A Study of the Research Direction and Trend in the Use of Corpus - Focusing on the Case of Japan -

말뭉치 구축·활용의 흐름과 현재의 동향 - 일본의 사례를 중심으로 -

In this paper, as a proposal to an effective corpus construction and utilization scheme, there is a purpose to explore the present situation of Japanese corpus construction, concrete content and current trend. In Japan, dependence on google is remarkable, there has been a steady effort to develop high-quality corpus and development tool. On the other hand, the Japanese corpus should clearly grasp the location and information from those created by individual researchers to their own purpose to those created mainly by universities, research institutes, national policy institutions, etc. It is difficult. In this survey, it was possible to distinguish by "media corpus", "literary·magazine·web and balanced corpus", "spoken language corpus", "learner corpus", "historical material corpus" etc. by field and type. In addition, there were not many tools developed for corpus efficient use and secondary processing such as "example search", "morphological analysis", "machine translation", etc. for tool corpus only. The current trend in Japanese corpus construction spurred preparations for Seed data which can be utilized in linguistic research and various fields of the fourth industry, including national policy and research institutes such as NINJAL, JPO, NICT, ALAGIN and companies such as RAKUTEN ing.