With the rapid advancement of multimedia technology and big data, the landscape of physical education (PE) research has undergone significant transformation. However, there remains a lack of quantitative and visual comparative analysis of PE research frontiers between China and the United States. This study adopts CiteSpace software to analyze 946 Englishlanguage publications from the Web of Science and 232 Chinese-language publications from the CSSCI database, constructing knowledge maps and clustering co-cited references to reveal research hotspots and trends over the past five years. Results indicate that US PE research primarily emphasizes health-oriented pedagogical models, teacher professional development, and evidence-based practices, whereas Chinese research focuses more on curriculum reform, teaching modes, and educational policy alignment. To enhance knowledge processing efficiency, an improved genetic algorithm combined with rough set theory (IGA+RS) is proposed for knowledge abbreviation. The algorithm introduces heuristic information on attribute significance into the genetic search process, integrates deletion, repair, and smoothing operators, and applies niche evolution to avoid premature convergence. Experimental results demonstrate that IGA+RS significantly reduces redundancy in decision tables while preserving classification accuracy, outperforming traditional rough set methods.
