Personalized Recommendation Mechanism of College English Writing Teaching Content Based on OBE Concept in Big Data Environment

With the rapid development of education information technology today, more and more educators are beginning to consider how to implement more personalized education and better align teaching to students' abilities. This paper explores personalized recommendations for college English writing instruction, utilizing the concept of Personalized Behavior Education (OBE). This paper first analyzed English-related data from 200 college students, including vocabulary, grammar, reading, and speaking abilities. Cluster analysis is used to classify the abilities of different groups in English writing. Collaborative filtering recommendation modeling is used to recommend writing methods for college students that are appropriate for them individually. This paper revealed that the first group of students utilizes 14% and 26% more quotations and advanced vocabulary, respectively, compared to the fifth group. Simulation experiments reveal that setting parameter 1 to 0.6 and parameter 2 to 0.2 results in a lower and more stable MAE value, thereby lessening its influence on the recommendation algorithm. Furthermore, the performance test results demonstrate that the precision value of this paper's scheme surpasses that of the traditional CF algorithm by a range of 0.088 to 0.14, and the recall value of this paper's algorithm surpasses that of the traditional CF method, with a maximum difference of 0.11. In addition, the actual accuracy of this paper's personalized recommender system is 6.2% higher than the expected one, and it takes only 1s to provide the user with a personalized recommender's response time. In summary, the experiments above reflect the high efficiency of this paper's recommender system and the high acceptance of students. © 2024 Xuemei Huang et al., published by Sciendo.