AI 진단 및 학습 추천용 학습 맵의 매핑 기능 개발을 위한 그래프 데이터베이스별 기능 분석

Abstract

In this study, the function of the graph database for efficient mapping function development was compared and analyzed in relation to the mapping function of the learning map (knowledge graph) for AI diagnosis and learning recommendation. Based on the commonalities and differences of functions by major graph database such as Neo4j, Microsoft Azure Cosmos DB, and Aerospike, five comparison criteria were set: data modeling, query language and API, data integrity and security, performance and horizontal scalability, community and support. Based on this, matters to be considered when selecting an appropriate graph database for the effective development of learning map mapping were derived. These considerations can be the basis for selecting an appropriate graph database when developing functions using a learning map. This study is expected to be used as a practical reference material when constructing a learning map using a graph database in the future.