Design of Individualized English Learning Path Recommendation Algorithm Based on Machine Learning and Internet of Things

Individualized English based on machine learning (ML) and Internet of Things (IoT) can play an important role in computer-aided instruction (CAI). This article introduces the problems existing when calculating the similarity of users and the importance of user comments to the recommendation system. Then, an emotion-based user similarity measurement method is proposed, which corrects the similarity between users by using the emotional comment information of users, thus improving the recommendation accuracy. In order to verify the effectiveness of this method, two methods are adopted to integrate user ratings and user comment information emotions, namely, the method based on opinion pre-filtering and the method based on user ratings embedding, and the recommendation results are evaluated. The experiment shows that the algorithm's functional index in collaborative filtering is higher than the same mean square error, and it is also verified that it is feasible to improve the recommendation effect by using auxiliary information such as user comments. This method improves the performance and accuracy of the recommendation model by integrating user ratings and user comments, and shows high stability in the face of different numbers of users evaluating projects together, thus ensuring the reliability of the recommendation system. © 2024, CAD Solutions, LLC. All rights reserved.