Zhengzhou Railway Vocational & Technical College, Zheng Zhou, Henan 450000, China.
School of Foreign Languages, Chuzhou University, Chuzhou 239000, Anhui, China.
Comput Intell Neurosci. 2022 Mar 10;2022:2709255. doi: 10.1155/2022/2709255. eCollection 2022.
As one of the most widely used languages in the world, English plays a vital role in the communication between China and the world. However, grammar learning in English is a difficult and long process for English learners. Especially in English writing, English learners will inevitably make various grammatical writing errors. Therefore, it is extremely important to develop a model for correcting various writing errors in English writing. This can not only be used for automatic inspection and proofreading of English texts but also enable students to achieve the purpose of autonomous practice. This paper constructs an English writing error correction model and applies it to the actual system to realize automatic checking and correction of writing errors in English composition. This paper uses the deep learning model of Seq2Seq_Attention model and transformer model to eliminate deep-level errors. Statistical learning is combined with deep learning and adopted a model integration method. The output of each model is sent to the n-gram language model for scoring, and the highest score is selected as output.
作为世界上使用最广泛的语言之一，英语在中国与世界的交流中起着至关重要的作用。然而，英语语法的学习对于英语学习者来说是一个困难而漫长的过程。特别是在英语写作中，英语学习者不可避免地会犯各种各样的语法写作错误。因此，开发一个用于纠正英语写作中各种写作错误的模型是非常重要的。这不仅可以用于英语文本的自动检查和校对，还可以使学生达到自主练习的目的。本文构建了一个英语写作错误纠正模型，并将其应用于实际系统中，实现了英语作文写作错误的自动检查和纠正。本文使用 Seq2Seq_Attention 模型和 transformer 模型的深度学习模型消除深层次的错误。结合统计学习和深度学习，采用模型集成方法。将每个模型的输出发送到 n 元语言模型进行评分，并选择得分最高的作为输出。
