용어지식의 통합 및 확장 방안 - 국립국어원 온용어 플랫폼을 중심으로 -

Abstract

This study presents a method for designing and utilizing a terminology knowledge platform that can contribute to the advancement of Korean AI technology by integrating and systematizing all the terminology knowledge built and utilized by various government agencies and organizations in the country. The study proposes a methodology for integrating databases of terminologies, such as glossaries, scattered across different agencies or fields into a unified system by applying a standardized classification system. It also discusses the process of linking terms with the same meaning and concept but presented in different forms across various glossaries, along with the associated issues and empirical solutions. Furthermore, the study explores ways to expand the terminology data into knowledge by adding semantic relationships and rules, and analyzes social phenomena by examining the statistical trends in terminology changes, thereby enhancing the usability of terminology knowledge. Furthermore, this study explores ways to address hallucination phenomena related to Korean language and culture in artificial intelligence through the construction of knowledge graphs. It also seeks development directions that can connect this effort with global knowledge networks.