A Study on the Creation of a Postpositional Particle Ontology in Developing an Old Korean Morpheme Dictionary

Morphological analysis of a document written in a natural language provides fundamental information for syntactic and semantic analysis.However, to this day, a high-performance Old Korean morpheme analyzer has not been developed.The main reason for this challenge is that constructing a comprehensive Old Korean morpheme dictionary remains a significant obstacle.This study aims to suggest a system that automates the creation of a postpositional particle ontology for an Old Korean morpheme dictionary by using a Modern Korean parallel corpus.Postpositional particles are morphemes that, while fewer in number than other morphemes, are used more frequently.Therefore, organizing the relationship between Modern Korean postpositional particles and Old Korean postpositional particles is essential for constructing an Old Korean morpheme dictionary.However, since a single Modern Korean postpositional particle can correspond to several different Old Korean forms depending on the era or work, manually searching for and registering them is an effort-intensive task.In this study, we propose a method for extracting postpositional particles from Modern Korean data in the Modern Korean-Old Korean parallel corpus and subsequently identifying and removing the corresponding postpositional particles from the Old Korean data.Additionally, we present a method for constructing a Modern Korean-Old Korean postpositional particle ontology that expresses the relationship between one Modern Korean postpositional particle and its various Old Korean forms.