This paper introduces the Adaptive Learning Path Navigation (ALPN) system, a
scalable approach for creating adaptive learning paths within E-learning
systems. The ALPN system employs an attention-based Knowledge Tracing (AKT)
model to evaluate students' knowledge states and a decision-making model using
Proximal Policy Optimization (PPO) to suggest customized learning materials.
The proposed system accommodates students' needs by considering personalization
parameters such as learning objectives, time constraints, and knowledge
backgrounds. Through an iterative process of recommendation and knowledge state
updating, the ALPN system produces highly adaptive learning paths. Experimental
results reveal the outstanding performance of the proposed system, providing
good insights into the future development of E-learning systems.

本文介绍了自适应学习路径导航（ALPN）系统，它使用基于注意力的知识跟踪（AKT）模型评估学生的知识状态，使用近端策略优化（PPO）的决策模型提供定制的学习材料，通过迭代的推荐和知识状态更新产生高度自适应的学习路径，实验结果表明，该系统的性能优异，为未来 E-learning 系统的发展提供了良好的启示。