With the increasing development of the digital economy and Internet technology, in order to adapt to the demand for teaching, many colleges and universities are changing their education mode through digital transformation. Starting from English teaching, this paper seeks to uncover a suitable way for the digital transformation of English education. Based on a collaborative filtering recommendation algorithm, this paper constructs an English teaching model that integrates personality recommendations. The model can use clustering algorithms to classify learners into different groups, which facilitates the subsequent implementation of personalized recommendations. The performance comparison and examples of English teaching in colleges and universities are used in this paper to analyze the efficient performance of the recommendation model and evaluate its usefulness in the smart classroom. The recommendation model designed in this paper divides learners into five clusters, which provides strong support for us to provide personalized recommendations of English learning resources in the future. This paper’s system, for example, in basic English, has a recommendation accuracy rate of 10% and 21% higher than reference systems 1 and 2, according to performance comparison. Applying this paper’s recommendation model in an efficient and intelligent English teaching classroom, after the experiment, the performance of class T improved by 4.29 points compared to class CK. To conclude, the recommendation system of this paper is very adaptable to the smart classroom and has a greater role in promoting the digital transformation of English teaching mode in colleges and universities.