從連結AI自然言語處理研究觀點來論多模式領域中之日本語研究未來展望

Future outlook of Japanese research in multimodal domain: From the viewpoint of connection to AI natural language processing research

本論文目的在於超越文學、語學、媒體等範疇形式，而針對現今流行表達方式之AI領域中言語研究表現特徵。主要是就日本語學、日本語教育之言語課題脈絡中來進行考案。AI之自然言語處理是不僅需要處理言語形式，也需要處理視覺資訊等視覺元素以及標記方式等非言語要素的多模式文本之研究領域。本論文將此特徵，放置於日本語表現研究史中，來考察並釐清日本語學、日本語教育之言語課題發展。

In this thesis, we will take up the language research in the AI; artificial intelligence as expression genre that is currently in fashion, and we would like to consider its expressive characteristics in the flow of language tasks of Japanese studies and Japanese language education beyond departmental genres such as literature, language, media, etc... In natural language processing of AI, it is necessary to handle multimodal text such as visual elements as visual information processing and non-verbal elements in notation etc., as well as language form. This is a field where researches are progressing now with methods and tasks different from Japanese language studies. In this thesis, we would like to consider these features as a development of language tasks on Japanese studies and Japanese language education from the view point of Japanese expression history.