Construction and Pedagogical Validation of an AI-Empowered Multidimensional Assessment System for English Speech Course

This study focuses on the in-depth integration of artificial intelligence (AI) technology with English speech evaluation, with the objective of developing a multidimensional, quantifiable, and visualizable intelligent evaluation system, and validating its effectiveness through practical teaching applications. The research is framed within the context of multimodal learning theory, incorporating four core dimensions—language, content, expression, and interaction—and explores the innovative application of AI in data collection, model development, and result feedback. Through comparative experiments and empirical analysis, the paper demonstrates the practical value of the intelligent evaluation system in enhancing students' speech capabilities, stimulating learning motivation, and optimizing teaching efficiency. The findings offer both theoretical underpinnings and practical guidance for the reform of foreign language teaching evaluation in the age of artificial intelligence.