Intelligent Matching and Optimization of Ideological and Political Elements in Computer Science Courses Based on AI Agent

Abstract

This study presents an AI Agent framework for intelligently matching and optimizing ideological and political elements in computer science (CS) curricula based on Outcome-Based Education (OBE) principles. The proposed system employs a hybrid recommendation approach, combining knowledge graph-based semantic reasoning (with 86+ validated ideological elements) and collaborative filtering (leveraging 320+ instructor profiles), achieving 40% higher matching accuracy (F1-score=0.79) than conventional methods. Key innovations include: (1) a three-dimensional taxonomy of CS ideological elements (ethical/legal, technological narratives, professional competencies), (2) an explainable recommendation engine providing justification paths (e.g., linking "sorting algorithms" to "engineering ethics"), and (3) continuous improvement mechanisms via real-time regulatory updates and longitudinal graduate tracking. Experimental results across five core CS courses demonstrate significant improvements in instructor satisfaction (4.1/5 vs 3.2/5 baseline) and student ideological awareness (+42% post-test scores). The framework addresses critical gaps in curriculum design by balancing technical rigor with value cultivation, offering a replicable model for STEM education.