Against the backdrop of deep integration between intelligent shipping and maritime data, the digital transformation of traditional maritime technology education has become an inevitable requirement for industry development. As the core resource documenting real-world vessel navigation behavior, Automatic Identification System (AIS) data holds irreplaceable value in analyzing collision avoidance decision-making logic and establishing objective quantitative teaching evaluation systems. This paper systematically reviews the research progress in the past five years, both domestically and internationally, on utilizing AIS data mining technology to identify collision avoidance scenarios, extract operational characteristics, and assess collision risks. It focuses on exploring application pathways of data analysis technologies in auxiliary maritime collision avoidance training, identifying core challenges and research gaps in current technical implementation scenarios. The study aims to provide theoretical foundations for vocational undergraduate programs to establish a "data-driven" collision avoidance teaching model, thereby fostering high-quality interdisciplinary maritime professionals equipped with data-driven thinking.