Research on Interactive Higher Education Quality Evaluation Model based on Affective Input Computing Framework

Abstract

This study presents an advanced framework for evaluating student engagement in higher education, leveraging the principles of affective computing. The framework, rooted in deep learning methodologies such as Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTM) networks, quantitatively analyzes students' vocal and facial expressions in real-time to assess emotional states. Experimental results validate the framework's efficacy in distinguishing between positive and negative emotions, offering educators a powerful tool for enhancing teaching strategies. By standardizing indicators and identifying correlations, clustering techniques are applied to develop an immediate assessment method for educational effectiveness. The framework's iterative design, which integrates feedback for continuous enhancement, ensures its relevance in evolving educational landscapes. This paper also addresses privacy concerns associated with data collection, outlining measures for safeguarding student information. Comparative analysis highlights the framework's superiority in scalability, cost-effectiveness, and privacy protection over conventional educational assessment methods.