Application of machine learning in intelligent transport systems: a comprehensive review and bibliometric analysis

The integration of machine learning (ML) into intelligent transport systems (ITS) has significantly enhanced the efficiency, security, and sustainability of urban transportation. Bibliometric analysis can identify existing research hotspots and the areas of interest and predict development trends, which can help to identify the current advancements in the role of ML in ITS. The Scopus database was used for the analysis because of its extensive collection of scholarly publications. This study examines a comprehensive set of 534 papers published between 2001 and 2023, utilizing various tools, including RStudio, VOSviewer, Excel, Publish & Perish, and Python to analyze data. The publication trend indicates that research in this area has grown exponentially each year. The keyword analysis indicated that key research interests within ML applications in ITS focus on network and communication technologies, urban mobility and infrastructure, safety and efficiency enhancements, and emerging topics such as demand-responsive transport and integrated decision-making. Europe leads in publication count, while East Asia, particularly China, contributes significantly, and research from regions such as North America, the Middle East, and Oceania, despite lower output, shows higher citation averages. This study also provides a comprehensive review of the top 50 most-cited publications, identifying six key focus areas and demonstrating the significant advancements in this field. However, this study is limited by keyword and tool biases, potential scope oversight, and reliance on a single database. Future research should expand keywords, use diverse tools, and include more databases. Overall, this study aims to serve as a valuable resource for researchers and practitioners, fostering innovation and developing more effective and sustainable transportation solutions.