短単位検索システムとOpenCHJ形式形態論情報付与ツールの構築

要旨

A web application, the "Short Unit Word Search System", that enables the searching of text data by short unit words, and a corresponding morphological annotation tool, "OpenCHJAnnotator", have been developed. Previously, searches based on short unit words primarily relied on "Chunagon", provided by the National Institute for Japanese Language and Linguistics. However, its use has been limited to specific corpora, making it difficult to utilize users’ own text data. To address this limitation, "OpenCHJAnnotator" was implemented to automatically convert users’ own text data into the OpenCHJ format, enabling the "Short Unit Word Search System" to process text data not included in "Chunagon". The integration of both tools enables seamless execution from data conversion to searching, while also providing verifiable reports of short unit word search results. Furthermore, generative AI models were utilized in the development of both tools, presenting a practical example of AI technology application in the field of language resources.

主題