学术英语阅读素材的智能选取与教学应用研究  

Data-driven source selection for academic English reading:An empirical study with pedagogical implications

作　　者：金檀[1] 李世炜 刘丰恺 JIN Tan;LI Shiwei;LIU Fengkai

机构地区：[1]中山大学外国语学院,广东广州510275 [2]中山大学物理学院,广东广州510275

出　　处：《外语界》2021年第1期20-27,共8页Foreign Language World

基　　金：国家社会科学基金一般项目“学术英语阅读的内容难度特征与智能模型研究”(编号18BYY110)的成果之一。

摘　　要：本研究基于学术作品的传播影响、在线读者的背景信息、文本的语言难度三大维度设计学术英语阅读素材智能选取方法,采用关注度分数、读者专业与职业背景、词汇覆盖率等量化指标对452篇学术报告的素材智能选取进行了实证探究。结果表明:第一,关注度分数能够有效筛选出传播影响力较强且涵盖较多学科专业的学术素材;第二,读者专业与职业背景数据能为特定素材的选取与使用提供真实可靠的读者档案;第三,词汇覆盖率能够有效控制学术素材的语言难度,并为分层教学提供决策参考。研究进而对智能选取方法的教学应用提供了可操作性建议。From three dimensions of the influence of academic works,background information of online readers,and linguistic complexity of texts,this study presents a methodological design of data-driven source selection for academic English reading.Using quantifiable indices of the Altmetric Attention Score,Mendeley reader profile,and Eng-Editor vocabulary coverage,the study carries out an empirical examination of the data-driven source selection from 452 research reports.The results are as follows:first,the Altmetric Attention Score can effectively screen out academic sources with high communication impact and covering a wide range of subjects and specialties;second,the Mendeley reader profile can provide authentic and reliable interpretation of readership for the selection and use of specific academic sources;third,the Eng-Editor vocabulary coverage can serve as the benchmark for linguistic complexity of academic sources,supporting decision-making for differentiation in teaching.The study further discusses pedagogical implications of the data-driven source selection.

关 键 词：学术英语阅读 素材智能选取 方法设计 实证研究 教学应用 

分 类 号：H319[语言文字—英语]