Scholarlink

<p>Real-time tracking and forecasting of student engagement and performance is the need of the hour in adaptive and smart learning environments especially in higher education. We introduce a Quantum-Inspired Multi-Modal Framework with three interconnected layers. The first layer commonly known as Quantum-Inspired Adaptive Signal Capture Layer which gathers multi-modal student signals and unifies it to create unique behavioral-cognitive fingerprints. The second layer- Cross-Perspective GNN (graph neural network) layer captures peer influence and tracks learning progression over time. The third layer- Hybrid Mind-Map &amp; Predictive Analytics Layer which dynamically assesses knowledge mastery and updates performance indicators thereby making the system adaptive and holistic. The proposed framework is evaluated against baseline models and existing works such as LSTM (long-short term memory), Transformer, and Spectral FNO (Fourier neural operator) based on metrics such as predictive accuracy (%), engagement correlation, and temporal adaptability (sec). Our framework achieved 97.6% accuracy, 0.91 engagement correlation (ρ), and 3.2 temporal adaptability thereby outperforming existing studies. The results show that the proposed framework improves predictions by focusing on the complete learning environment factors like peer progress and relative performance scores for personalized learning.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Quantum-inspired multi-modal framework for real-time student engagement and performance prediction in smart learning environment

  • 期刊论文

  • 发表日期：

    2025年10月25日

查看全文
  • Shirly Abraham,
  • Deepa Pillai,
  • Trupti Bhosale

摘要

Real-time tracking and forecasting of student engagement and performance is the need of the hour in adaptive and smart learning environments especially in higher education. We introduce a Quantum-Inspired Multi-Modal Framework with three interconnected layers. The first layer commonly known as Quantum-Inspired Adaptive Signal Capture Layer which gathers multi-modal student signals and unifies it to create unique behavioral-cognitive fingerprints. The second layer- Cross-Perspective GNN (graph neural network) layer captures peer influence and tracks learning progression over time. The third layer- Hybrid Mind-Map & Predictive Analytics Layer which dynamically assesses knowledge mastery and updates performance indicators thereby making the system adaptive and holistic. The proposed framework is evaluated against baseline models and existing works such as LSTM (long-short term memory), Transformer, and Spectral FNO (Fourier neural operator) based on metrics such as predictive accuracy (%), engagement correlation, and temporal adaptability (sec). Our framework achieved 97.6% accuracy, 0.91 engagement correlation (ρ), and 3.2 temporal adaptability thereby outperforming existing studies. The results show that the proposed framework improves predictions by focusing on the complete learning environment factors like peer progress and relative performance scores for personalized learning.