摘要
本文是基于BP人工神经网络的"iWrite 2.0"英语写作教学与评阅系统的语用研究。首先追溯了以PEG、IEA和E-rater为表性的三大国外自动作文评阅系统理论依据的形成背景;然后比较iWrite与国内"冰果"和"句酷"评阅系统之间的异同特点,凸显自身两大特征:"联想词库"内容评价法和GEC"错误类型"识别法;接着提出基于BP人工神经网络的新技术;最后阐释了该系统测评维度上"重语义,轻语用"的弊端,并给与有效建议。
The paper is a pragmatic study on iWrite English writing teaching and evaluation system with BP artificial neural network,on the basis of theoretical comparison from three representative automated essay grading systems including PEG、IEA和E-rate,which compares iWrite with Bingguo and Juku of domestic grading systems.With the characters of Associative Thesaurus and Grammar Error Correction,it discusses BP artificial neural network and provides a pragmatic proposal.
作者
武喆
WU Zhe(International Business School,Dalian University of Finance and Economics,Dalian Liaoning 116622)
出处
《山东农业工程学院学报》
2020年第12期44-48,共5页
The Journal of Shandong Agriculture and Engineering University
基金
辽宁省高校外语教学科研项目“基于iWrite(英语写作教学与评阅系统)对大学英语四六级写作教学改革的研究”
全国高校外语教学科研项目“基于智能评阅的大学英语四六级写作教学改革研究”(2019LN0044A)。
