Design and Construction of a Knowledge Database for Learning Japanese Grammar Using Natural Language Processing and Machine Learning Techniques

Design and Construction of a Knowledge Database for Learning Japanese Grammar Using Natural Language Processing and Machine Learning Techniques

연구 분야: Artificial Intelligence

학회: 2022 4th International Conference on Natural Language Processing (ICNLP)

This work describes the design and construction of a computer-aided foreign language learning grammar knowledge database for learning Japanese sentence patterns using natural language processing (NLP) and machine learning techniques. This grammar knowledge database can be applied for learners of Japanese as a second language (JSL) to study the Japanese sentence patterns for Japanese Language Proficiency Test (JLPT). Selection of Example sentence is a challenging problem when constructing the database, because it is difficult to detect the Japanese sentence patterns from the example sentences. To solve this problem, NLP and machine learning technologies are applied to automatically detect the Japanese sentence patterns from the input example sentences by proposing an approach which integrates Conditional Random Fields (CRF) machine learning algorithm and manual rules. Several experiments were conducted and the experimental results demonstrated the validity of the proposed methods by using NLP and machine learning techniques on the design and construction of the Japanese grammar knowledge database. This database can be considered as an effective supplementary tool for learning Japanese grammar, and as a further improvement in the effectiveness of current intelligent computer-aided Japanese language learning system.