Strategies for Enhancing Student Engagement in Mobile Learning Environments through Interactive Technologies

Abstract

The widespread adoption of mobile Internet technologies has established mobile learning as a prominent educational paradigm. However, the inherent characteristics of fragmentation and weak supervision in mobile learning present significant challenges in maintaining sustained student engagement. Effective strategies for dynamically monitoring and enhancing engagement are critical for optimizing the outcomes of mobile learning environments. While existing research has focused on personalized resource recommendations, common issues such as data sparsity and the cold-start problem persist. Moreover, traditional recommendation models fail to capture the complex, nonlinear interactions between students and resources, thereby limiting the precision of engagement-driven resource allocation. To address these shortcomings, an interactive resource recommendation algorithm targeting student engagement in mobile learning is proposed. Engagement levels are quantified through sentiment analysis using a Naïve Bayes classifier, which informs a deep integration of a dual-tower neural network and matrix factorization model. This approach utilizes nonlinear mapping to uncover deep semantic relationships between students and resources, enabling the precise identification of resources that can enhance student engagement. Building on this, a systematic strategy for improving student engagement is introduced, encompassing real-time emotional interventions and personalized learning path generation. This strategy creates a closed-loop system that spans from engagement perception to targeted intervention. The proposed methodology innovatively integrates sentiment analysis as a dynamic metric for engagement and combines it with deep matrix factorization models, thereby overcoming the limitations of traditional recommendation frameworks that rely on explicit ratings and linear assumptions. Furthermore, a dual-layer strategy is proposed, combining real-time interventions with personalized path generation, offering a comprehensive engagement enhancement plan for mobile learning platforms. This approach provides a full-cycle solution, from immediate response to long-term strategic planning, which holds significant theoretical and practical value for advancing mobile learning systems.