With the development of technologies such as big data analysis and deep learning, various industries have begun to integrate with big data analysis and deep learning and continue to promote the development of the industry. This system is an intelligent writing scoring system for college English teaching. It uses popular big data analysis and deep learning to distinguish training algorithms. From 2015 to 2022, the number of college students taking exams has increased yearly, with an increase of more than 50%. Therefore, the system proposes a text vector calculation method that can find matching samples in the text set after the text is weighted by the weight function and uses deep learning to distinguish the algorithm evaluates the matched text, and finally can get the final score according to the content quality, semantic coherence, text readability, and other aspects of the text. Compared with traditional manual scoring, this technology is more convenient, quick, concise, and effective. This system is significant for improving the efficiency of teaching English writing in college.