특허 토픽 기반 메타버스 기술 프레임워크: 가상 세계를 중심으로

This paper aims to develop a metaverse technology framework using patent data and analyze the capabilities of companies based on it. To this end, 4,099 patents registered with the US Patent and Trademark Office were collected and preprocessed. Latent Dirichlet Allocation(LDA) topic modeling was performed to derive 20 topics, and a 4-layer metaverse technology framework was derived by grouping these topics. Leading companies were identified and their capabilities analyzed from the result of LDA topic modeling. This study provides a data-based empirical perspective to understand the technology structure and hierarchy of metaverse service; it can be used for setting companies" future metaverse research plans and standardization activities.