BERTopic을 이용한 교통 제어 시스템 AI 기술 동향 분석 = Trend Analysis of Traffic Control Systems Using Topic Modeling

다국어 초록 (Multilingual Abstract)

Purpose This study aims to identify trends in AI technologies for traffic control systems, analyze future technology trends through patent data analysis, and propose future development directions based on these insights.
Methods A total of 5,453 patents related to AI technologies in traffic control systems were collected. The BERTopic modeling technique was applied to analyze key subfields and core keywords. Additionally, the changes in key concepts and keywords over time were visualized and systematically analyzed to identify distinct AI technology areas within traffic control systems.
Results The analysis identified a total of nine distinct topics within the field of AI technologies for traffic control systems. Specifically, two topics emerged during the early adoption phase (1990–1999), three during the growth phase (2000–2013), and eight during the expansion and maturity phase (2014–2023). AI technology application was notably prominent in Topic 0 (AI-based image analysis and video-based traffic control) and Topic 1 (traffic flow prediction and neural network-based signal control).
Conclusion Utilizing BERTopic to analyze patent data, this study provides meaningful insights into trends in AI technology for traffic control systems. By detailing how technological innovation is advancing across specific subfields, the findings offer foundational information useful for establishing strategic directions for research and development within the AI traffic control domain.