101. Research on Error Detection Technology of English Writing Based on Recurrent Neural Network
- Author
-
Li Yong-An, Ma Lun, Wang Wei, and Qu Qianqian
- Subjects
Grammar ,business.industry ,Computer science ,media_common.quotation_subject ,Feature extraction ,computer.software_genre ,Grammatical error ,Information extraction ,Recurrent neural network ,Reading (process) ,Active listening ,Artificial intelligence ,business ,Error detection and correction ,computer ,Natural language processing ,media_common - Abstract
In learning English, learners are faced with various challenges, such as listening, speaking, reading and writing. Among them, writing is the most important and difficult one, and grammatical errors are the most common type of errors in English writing. The research and implementation of grammatical error detection and correction in English writing is of great significance to both English learners and English teachers. In this paper, we propose a sequence annotation model based on recurrent neural network to solve the problem of the influence of grammar errors and other noises in English writing corpus on sequence information extraction. Aiming at the grammatical errors in English writing, this paper proposes two methods of English grammatical error detection and correction. The experimental results show that the accuracy of RNN is higher than that of CAMB after 10 to 20 iterations, ranging from 0.848 to 0.916.
- Published
- 2021