With the continuous development of data integration ability, process acquisition means and intelligent analysis methods of education management system, the implementation of traditional national physical education curriculum has the conditions of digital modeling and dynamic analysis. Focusing on the application requirements of artificial intelligence to support curriculum implementation regulation in education management system, this paper constructs an intelligent analysis framework that integrates curriculum intelligent modeling, implementation state analysis, strategy generation and feedback correction. The curriculum resources, teaching progress, student participation behavior, action performance and process evaluation records are unified coded, correlated represented and calculated collaboratively. Based on 3120 curriculum implementation samples, 18640 platform logs, 9240 classroom interaction records and 5280 sets of action recognition results from three higher vocational colleges, the study uses graph structure representation, timing state coding, strategy probability allocation and rule constraint decision methods to complete curriculum implementation analysis and strategy output. The experimental results show that the accuracy of course implementation matching reaches 92.6%, the consistency rate of strategy recommendation reaches 89.8%, and the average response delay of the system is controlled at 1.7 seconds. The results show that the framework can enhance the precision of curriculum organization adaptation and the stability of feedback correction while maintaining good response efficiency, and provide computational support, application basis and practical reference for the intelligent implementation of traditional national physical education curriculum in education management system.
Povzetek: Za potrebe regulacije izvajanja tradicionalnih etničnih športnih predmetov ta članek v okviru izobraževalnega upravljavskega sistema vzpostavlja inteligentni analitični okvir za skupno modeliranje učnih virov, učnega napredka, vedenja udeležencev, gibalne izvedbe in zapisov procesnega vrednotenja. Na podlagi 3120 vzorcev izvajanja pouka ter povezanih podatkov s platforme in iz učnega procesa rezultati testiranja kažejo, da natančnost ujemanja izvajanja pouka dosega 92,6 %, stopnja skladnosti priporočanja strategij dosega 89,8 %, povprečna odzivna zakasnitev sistema pa znaša 1,7 sekunde, pri čemer celotno delovanje ostaja stabilno.
