DNN-Based Resource Recommendation for Ideology Theory Courses Online

Abstract

Revamped IPE confronts static material constraints and outdated pedagogy, warranting integration of web resources and big data analytics for instructional innovation. Digital IPE adoption in vocational education optimizes online resource use, enhancing teaching effectiveness. Introducing CUPMF, a personalized learning model, we conduct empirical assessments on a large dataset (364,617+ entries) from Smart Classroom's cloud platform and public datasets, reflecting varied IPE scenarios. Comparative experiments against association rule, content-, tag-based, and collaborative filtering algorithms show CUPMF's superiority. It achieves a 11.61% F1 score boost over four alternatives for basic recommendations and outperforms Que Rec by 1.975%. Complexity-wise, CUPMF registers an 11.52% mean F1 score increment over four methods and 1.875% over Que Rec. Proven, CUPMF markedly improves IPE resource recommendation accuracy and efficacy, poised to transform personalized online vocational learning.