基于Transformer模型的中文文本自动校对研究  

Research on automatic proofreading of Chinese text based on Transformer model

作　　者：龚永罡 裴晨晨 廉小亲 王嘉欣 Gong Yonggang;Pei Chenchen;Lian Xiaoqin;Wang Jiaxin(College of Computer and Information Engineering Beijing Key Laboratory of Food Safety Big Data Technology,Beijing Technology and Business University,Beijing 100048,China)

机构地区：[1]北京工商大学计算机与信息工程学院食品安全大数据技术北京市重点实验室

出　　处：《电子技术应用》2020年第1期30-33,38,共5页Application of Electronic Technique

摘　　要：提出将Transformer模型应用于中文文本自动校对领域。Transformer模型与传统的基于概率、统计、规则或引入BiLSTM的Seq2Seq模型不同,该深度学习模型通过对Seq2Seq模型进行整体结构改进,从而实现中文文本自动校对。通过使用公开数据集对不同模型进行对比实验,采用准确率、召回率与F1值作为评价指标,实验结果表明,Transformer模型相比较于其他模型,在中文文本自动校对的性能上有了大幅提升。This paper proposes to apply Transformer model in the field of Chinese text automatic proofreading. Transformer model is different from traditional Seq2 Seq model based on probability, statistics, rules or BiLSTM. This deep learning model improves the overall structure of Seq2 Seq model to achieve automatic proofreading of Chinese text. By comparing different models with public da-ta sets and using accuracy, recall rate and F1 value as evaluation indexes, the experimental results show that Transformer model has greatly improved proofreading performance compared with other models.

关 键 词：中文文本校对 Transformer模型 深度学习 

分 类 号：TN957.52[电子电信—信号与信息处理] TP312[电子电信—信息与通信工程]