This research employs big data analysis technology to investigate the learning behavior patterns of international students and examines the correlation between these patterns and academic performance. By conducting cluster analysis on data from 1,000 international students using the Coursera platform, we discovered three primary learning behavior patterns: efficient learners, average learners, and inefficient learners. The research results show that efficient learners are outstanding in terms of online learning time and forum activity, and their academic performance is significantly higher than other categories of students. This discovery corroborates the beneficial connection between active participation in learning and achievement in academics, offering evidence to guide educators in course design and teaching methods. Furthermore, the analytical model developed in this research not only improves comprehension of international students’ behavioral patterns but also offers data-driven decision-making aid for educational institutions, showcasing the potential of big data in the education sector.