In this paper, firstly, based on cluster analysis in data mining technology, we propose to build an English online learning platform based on the system of B/S structure and realize the research on the innovation path of the English culture teaching method through the online learning platform. Secondly, the clustering analysis is carried out by taking a certain number of students’ level test data from the database of the studied English culture online learning platform, which is divided into four modules: basic English, professional English, business English and English listening, and the process of grouping or classifying the given set of data samples by using K-means algorithm, while the data attributes can be calculated according to the objective function correlation. Then, three stages of English culture teaching platform before, during and after class were selected as the research objects, and the K-means algorithm was used to analyze the correlation of English teaching platform satisfaction. The results show that there is a positive correlation between satisfaction of the English culture teaching platform before, during and after the lesson and the correlation strength is |rs| > 0.79, P = 0.000 < 0.01, which is highly correlated. Among them, the correlation coefficient between pre-class and in-class reached 0.871. This study creates an information-based teaching environment and innovates English teaching methods to improve students’ comprehensive English literacy.