期刊文献+

面向SELL语料库的AI虚拟英语口语训练系统研究 被引量:2

Study on AI Virtual English Speaking Training System for SELL Corpus
下载PDF
导出
摘要 为了改善英语学习方式中的口语学习环境和效率,基于AI虚拟现实技术,建立了适用于中国英语语言特征的SELL语料库。从VR交互界面、语音识别和智能对话3方面对系统设计进行详细阐述,最后选取某高校两组学生班级进行了一个学期的英语学习对比试验。实践发现:首先,虚拟现实英语学习系统逻辑清晰,基本实现了课程个性化定制和三维虚拟操作,并且支持外部VR设备,为英语学习者提供了沉浸式的英语学习平台;其次,基于Kaldi语音识别工具训练后,修正的SELL语料库测试集的准确率基本上可以控制在12.1%,验证集准确率控制在11.5%;最后,对比英语学习实践发现试验组在听力和口语表达方面成绩有显著提高,口语表达方面平均成绩较之前提高了19.6%,听力得分较之前提高13.3%,试验结果表明该智能化虚拟现实学习训练系统效果较好。 In order to improve the traditional study way of spoken English learning environment,AI is established first in China SELL corpus of English learning environment,and then from the VR interface design,design of speech recognition system and intelligent dialogue system are introduced in detail,finally by choosing a university to form two groups of students,the contrast test is carried out in one semester of English learning for evaluation.The research finds that first,the virtual reality English learning system has clear logic,basically realizes course customization and 3 D virtual operation,and supports external VR equipment,providing an immersive English learning platform for English learners.Secondly,the accuracy of the modified SELL corpus test set can be basically controlled at 12.1%,while the accuracy of the verification set is controlled at 11.5%.Finally,it is found that the test group has a significant improvement in listening and oral expression.The average score of oral expression is 19.6%higher than before,and the listening score is 13.3%higher than before.The test results show that the intelligent virtual reality learning and training system has a better effect.
作者 林辉 LIN Hui(Foreign Language Teaching and Research Department,Shanxi College of Communication Technology,Xi’an,Shanxi 710018,China)
出处 《微型电脑应用》 2020年第7期126-129,共4页 Microcomputer Applications
关键词 SELL语料库 虚拟现实 英语口语 SELL corpus virtual reality oral English
  • 相关文献

参考文献12

二级参考文献93

共引文献75

同被引文献10

引证文献2

二级引证文献2

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部