AttentiveLearn: Personalized Post-Lecture Support for Gaze-Aware Immersive Learning

Paper Title

AttentiveLearn: Personalized Post-Lecture Support for Gaze-Aware Immersive Learning

Publication Info

• Topic area: Immersive learning systems and post-lecture personalized support.
• Keywords: Immersive learning, VR classrooms, attention-aware systems, personalized quizzes, eye-tracking, mobile learning assistant, learner engagement, motivation, learning outcomes.

Background and Problem

• Problem / challenge: While immersive virtual classrooms in VR have shown promise for enhancing learning experiences, most research focuses on in-lecture support. Post-lecture support, which is critical for sustained motivation, engagement, and learning outcomes, remains underexplored, especially in immersive contexts.
• Significance: Addressing post-lecture support gaps can improve learner-centered designs, enabling better attention management, motivation, and engagement beyond the lecture session.
• Motivation and related work: Prior work has demonstrated the effectiveness of attention-aware systems and personalized quizzes in non-immersive settings. However, immersive learning lacks frameworks to transfer in-situ attention monitoring into ex-situ personalized support. This paper builds on research into gaze-based attention metrics and adaptive quizzes to bridge this gap.

Solution

• Proposed approach: AttentiveLearn, an attention-aware learning ecosystem that uses eye-tracking data from VR lectures to generate personalized post-lecture quizzes delivered via a mobile learning assistant.
• Novelty:
  1. Conceptualization of a framework bridging in-situ immersive learning with ex-situ personalized support through attention-aware quizzes.
  2. Integration of an attention-aware personalization pipeline into a mobile assistant, demonstrating feasibility in real-world educational settings.
  3. Empirical insights from a four-week field study investigating effects on motivation, engagement, and learning outcomes.
• Procedure and key techniques:
  • Eye-tracking data collected during VR lectures are processed into attention metrics (AOI coverage, attention switches, Attention Distribution Index).
  • Metrics inform personalized quiz generation targeting low-attention sections.
  • Quizzes are delivered via a mobile assistant with additional features like Q&A (OpenChat) and follow-up practice quizzes (ChatQuiz).

Results

• Concrete findings:
  • Attention-aware quizzes were rated as significantly more accurate in reflecting attention gaps (attentive group: M = 4.21, non-attentive group: M = 2.94, p = 0.003).
  • Participants in the attentive group reported higher engagement (ω2 = 0.09) and motivation, particularly in intrinsic goal orientation (ω2 = 0.177), task value (ω2 = 0.137), and self-efficacy (ω2 = 0.070).
  • Intermediate Mini-Exam scores remained stable for the attentive group but declined for the non-attentive group in week 3 (p = 0.004).
• Advantage over baselines:
  • Attention-aware quizzes revealed gaps more effectively than non-personalized quizzes, improving engagement and motivation.
  • AttentiveLearn buffered against performance decline in intermediate assessments under increased difficulty and external pressure.
• Experiments / evaluation:
  • Four-week field study with 36 university students attending VR lectures on Bayesian data analysis.
  • Mixed-methods evaluation combining eye-tracking data, quiz scores, surveys (UES-SF, MSLQ), and interviews.
  • Between-subjects design comparing attentive and non-attentive groups.
• Limitations and future work:
  • Small sample size (n = 36) and limited diversity restrict generalizability.
  • Study focused on one lecture topic; broader application across domains is needed.
  • Attention metrics rely on eye-tracking, which may not fully capture cognitive engagement. Future work should integrate multimodal signals and user-reported focus.

Summary

AttentiveLearn bridges immersive VR lectures with personalized post-lecture support, leveraging eye-tracking data to generate quizzes targeting attention gaps. A four-week field study demonstrated improved engagement, motivation, and intermediate learning outcomes for students using attention-aware personalization. While the system did not significantly impact final exam scores, it showed promise for sustaining learning progress under pressure. Future research should explore multimodal personalization, diverse learning domains, and extended support features to enhance immersive learning ecosystems.