Abstract：To solve the problem that the current recommendation methods could not fully consider the learning objectives of learners and the dynamic uncertainty of learner′s cognitive level, a knowledge mesh recommendation method was proposed based on fuzzy rules. In the proposed method, the teaching experience of the course teacher and the learning path of the learners who had learned the course were respectively formed into each knowledge mesh. The cognitive level of the learners was determined by comprehensively considering the factors of questionmaking time, test difficulty and memory. The similarity between the knowledge meshes was calculated by fusing the matching degree and cognitive level, and the knowledge meshes were clustered by the fuzzy Cmean clustering method. Taking the learning paths of 15 learners as experimental data, the knowledge meshes of W1-W15 were constructed, and the fuzzy cognitive level and membership degree of W1-W15 were calculated. The knowledge mesh with the largest membership degree was optimized and combined into a new knowledge network. The experimental results show that the proposed recommendation method can recommend suitable learning paths for learners′ learning purpose and reasonably reflect learners′ cognitive process.
