Research on multi-dimensional evaluation method of sports interest of college students

Abstract

This study investigates students’ engagement with sports content, employing a novel method to quantify interest levels through observable physical responses. By analyzing facial expressions and body language, specifically the muscle activities across various body parts, we derive a unique numerical value to represent the intensity of pleasure experienced by viewers in each frame. This approach involves calculating the rotational angle weights between limbs and their movements over consecutive frames, leading to a comprehensive evaluation score for the observed actions. Utilizing the Openpose platform for data analysis, we discovered a significant correlation (r=0.43) between real-time assessment scores and the “sports participation” interest dimension. Analysis of facial expressions in educational sports videos revealed “concentration” as the predominant expression, indicating high student interest. Furthermore, the study identified instances of interest fatigue within an 18-minute duration, notably between frames 32005 and 36507, with a PERCLO value of 0.46, demonstrating the method’s potential in assessing and enhancing engagement with sports education.