Abstract
As artificial intelligence (AI) continues to evolve, its applications in the realm of education are increasingly becoming prevalent. English composition (EC), a crucial language expression skill, serves as a vital medium for students to engage in input and output activities in their second language, garnering significant attention from educators and second language researchers. By harnessing AI to create an intelligent essay correction system, establish a personalized learning platform, and facilitate real-time interaction and communication, we can significantly enhance the efficiency and quality of English essay teaching, thereby fostering the comprehensive development of students. Drawing upon these principles, this article introduces an AI-assisted deep learning (DL) based EC error automatic correction system, encompassing a front-end learning unit and a back-end support unit. This system preprocesses collected EC fragments, extracts features using inter-segment distance partitioning factors, and employs dynamic time normalization techniques to correct and grade identified EC errors. The system then relays the recognition, correction, and scoring results to users through a feedback module. Simulation results demonstrate that the system described in this article effectively supports teachers in efficiently conducting essay corrections and data analysis, addresses the challenge of limited teacher correction and evaluation, boosts teaching efficiency, and spurs student initiative in writing.