基于BERT的汉语作文自动评分研究  

Research on Automatic Scoring of Chinese Composition Based on BERT Model

作　　者：李琳[1] 董璐璐 马洪超[2] LI Lin;DONG Lulu;MA Hongchao(Qinghai Normal University,Xining 810016,China;Beijing Language and Culture University,Beijing 100083,China)

机构地区：[1]青海师范大学计算机学院,西宁810016 [2]北京语言大学发展规划与学科建设办公室,北京100083

出　　处：《中国考试》2022年第5期73-80,共8页journal of China Examinations

基　　金：汉考国际2020年度科研基金项目“基于深度学习方法的作文自动评分研究”(CTI2020B05)。

摘　　要：双向编码表示转换模型(bidirectional encoder representations from transformers,BERT)是基于Transformer架构的预训练语言模型,解决了经典神经网络的数据依赖和语言编码问题,在自然语言处理领域取得很大成功。以汉语水平考试(HSK)作文为对象,对BERT模型进行迁移和微调,构建基于BERT的汉语作文自动评分模型。结果表明:对模型的网络结构、池化策略和学习率进行优化后,评分效果最好,模型预测出的分数与人工评分结果具有较强一致性。BERT(bidirectional encoder representations from transformers)is a pre-trained language model based on Transformer architecture.It solves the problems of data dependence and language coding that happen to classical neural networks and has achieved great success in the field of natural language processing.Taking the composition part of the Chinese Proficiency Test(or HSK)as the object,this paper constructs an automatic scoring model of Chinese composition based on BERT while transferring and fine-tuning the BERT model.The results show that after optimizing the network structure,pooling strategy and learning rate of the model,the scoring works the best and the scores predicted by the model are in good agreement with the manual scores.

关 键 词：作文自动评分 自然语言处理 预训练语言模型 BERT 深度学习 

分 类 号：G405[文化科学—教育学原理]