Enhancing the pedagogical model of ideological and political education represents a critical endeavor in contemporary educational institutions. With technological advancements, particularly in artificial intelligence, educational platforms encounter formidable challenges. This study proposes an advanced intelligent teaching platform tailored for ideological and political education at higher education institutions, constructed on the foundational principles of consistency, realism, human-centricity, purposefulness, operational feasibility, and systematic integration. This platform innovatively incorporates the Bidirectional Memory Enhancement Network (BMEN), which integrates an attention mechanism responsive to the unique memory and forgetting patterns of students. Consequently, a novel Knowledge Tracking Method Based on Bidirectional Memory Enhancement Attention (BMEAKT) is developed to precisely gauge students' comprehension levels in ideological and political subjects. Empirical results indicate that this platform surpasses traditional baseline models in tracking knowledge, exhibiting General Area Under the Curve Metric (GAUCM) values between 0.6549 and 0.7234, thereby effectively assessing students' mastery over the subject matter. A visual case study further validates the platform's utility, demonstrating a 59.12% accuracy in predicting students' correct responses to exercises. This research underscores a viable strategy for profoundly enhancing the instruction of ideological and political education in tertiary education settings. © 2024 Xinhong Feng., published by Sciendo.
