The growing global interest to decarbonize the transportation industry has resulted in numerous scientific publications. This study reviews the rapidly expanding body of research and identifies the knowledge gaps in transport decarbonization between regions. This study employs a hybrid approach combining topic modeling and machine learning to identify research topics and their knowledge structures, and then compares the main debated topics between Asia and non-Asian regions. A dataset of 777 articles, including 410 Asian and 367 non-Asian articles, published between 1990 and 2022 was extracted from the Scopus database. The latent Dirichlet allocation topic modeling results showed that five potential topics were derived from Asia, while six were derived from non-Asian regions, and the knowledge structure of each topic differed between the two regions. The K-nearest neighbor machine learning algorithm results indicated a 92% accuracy for Asian topics and an 89% accuracy for non-Asian topics. The findings suggest that the Asian studies focused on “energy use in transportation” and “drivers of CO2 emissions in transportation,” while the non-Asian studies focused on “electric vehicles” and “fuel consumption.” This paper will keep academics and practitioners updated on the paradigm shift in the research on transportation decarbonization.

[Key Words] Decarbonization, Transportation, Asia, Topic Modeling, Machine Learning