Context-aware sequential course recommendation via conditional variational autoencoders and transformer architectures...

Area of Science:

• Artificial Intelligence
• Educational Technology
• Machine Learning

Background:

• Personalized course recommendations are challenging in dynamic e-learning platforms with fluctuating student preferences.
• Traditional algorithms struggle with static or minimally contextualized user behavior, limiting effectiveness in diverse learning settings.

Purpose of the Study:

• To develop a context-sensitive sequential recommendation framework for e-learning.
• To address the limitations of traditional algorithms in dynamic and IoT-influenced educational environments.

Main Methods:

• Utilized Conditional Variational Autoencoders for robust learner and course representations from sparse data.
• Employed a Transformer-based sequential model to capture long-term learning dependencies.
• Integrated contextual signals from IoT-enabled environments into the recommendation process.

Main Results:

• Demonstrated statistically significant improvements in ranking and error metrics compared to baseline methods.
• Validated the efficacy of uncertainty-aware modeling and context-driven sequential learning.
• Showcased consistent personalization across diverse learner profiles.

Conclusions:

• Reinterprets course recommendation as a dynamic, context-sensitive process.
• Establishes a scalable framework for intelligent e-learning systems.
• Facilitates personalized learning trajectories within IoT-integrated digital education environments.