摘要
第二语言教学和自然语言处理以往从研究范式上看差异明显,背后反映出的则是基于"规则"和基于"统计"的不同的研究思路。自然语言处理经历了由偏重"规则"到重视"统计"的发展,这对第二语言教学的发展具有重要的启示作用。从统计视角看,为分别解决人和计算机"学习"语言的问题,第二语言教学和自然语言处理的研究思路本质上应当是相通的。其相通之处在于,自然语言处理是以统计语言模型来模拟人在语言学习和使用中的认知特点,而基于统计的自然语言处理实践又有助于揭示出语言教学的规律,即:不应过高估计语法在语言学习中的作用,要使学习者建立正确的语言模型,必须依靠大量规范的语料的训练。
The difference between the second language teaching and Natural Language Processing’s previous research paradigm is obvious,which reflects the different research ideas based on“rules”and“statistics”.Natural Language Processing has experienced the development from the emphasis on“rules”to“statistics”,which has an important inspiration for the development of second language teaching.From a statistical perspective,in order to solve the problem of“learning”language separately,the ideas of the second language teaching and Natural Language Processing’s research are similar.The similarity is that Natural Language Processing simulates the human’s cognitive characteristics in language learning and use with the statistical language model,and the practice of Natural Language Processing based on statistics helps to reveal the laws of language teaching.The laws are as follows:the role of grammar in language learning should not be overestimated;in order to establish the correct language model for the learners,it is necessary to train them with a large number of standardized linguistic data.
出处
《国际汉语教育（中英文）》
2018年第1期39-45,共7页
International Chinese Language Education
关键词
第二语言教学
自然语言处理
统计
本质
second language teaching
Natural Language Processing
statistics
essence
