E-learning has increased in popularity, especially during the COVID-19, due to its numerous advantages that allow learners to study anywhere and anytime. Therefore, recommending a list of the most appropriate learning objects for learners according to their specific needs is a great challenge for adaptive e-learning systems. In an e-learning environment, the optimum adaptive e-learning system is one that can adapt dynamically to the profile of each learner. Within that particular context, various approaches were proposed. In this article, we propose a new adaptation technique based on learner's learning style and motivation score by using collaborative filtering technique, constrained Pearson correlation coefficient, adjusted cosine measure, and K-nearest neighbor algorithms. The proposed approach is focused on how to develop and construct an effective customized pedagogical learning scenario for learning resources, and improve the accuracy of the adaptation by choosing the most suitable learning objects for learners. Therefore, we used the dataset MovieLens100K containing 943 learners and 1,682 learning objects. Additionally, a few experiments have been conducted to validate the performance of our technique. The results indicate that taking into account the learner's learning style and motivation score can completely satisfy the customized needs of learners and improves the quality of learning.