E-learning improves the shareability and reusability of knowledge, and surpasses the constraints of time and space to achieve remote asynchronous learning. Since the depth of learning content often varies, it is thus often difficult to adjust materials based on the individual levels of learners. Therefore, this study develops an ontology-based adaptive dynamic knowledge concept e-learning mechanism that generates learning maps based on learner characteristics and guides learners effectively. To achieve this goal, this study proposes an adaptive dynamic concept e-learning navigation procedure, designs learning models based on the adaptive learning needs of learners, and develops knowledge map model and learning map model. Finally, this study designs adaptive dynamic concept learning map-planning algorithms based on the particle swarm optimization (PSO) algorithm. The learning maps generated by these algorithms meet the dynamic needs of learners by continually adjusting and modifying the learning map throughout the learning process. Adapting the adaptive learning content according to the dynamic needs of learners allows learners to receive more instruction in a limited period.