学科教育研究中的机器学习:分析框架、国际比较与启示

Machine Learning in Discipline Education Research:Analytical Framework,International Comparison and Implications

机器学习在学科教育中具有巨大应用潜力。文章通过对CNKI及Web of Science数据库中近五年文献的梳理,构建分析框架比较国内外应用模式的异同:国外精于技术追求,国内重在落实教学服务初衷;国外重视课堂内外的学生管理,国内关注学习习惯养成;国外专注思维动机内部情感挖掘,国内聚焦表情反应外部情感提取;国外提倡媒体覆盖课内外辅导管理,国内推广大规模在线教育应用;国外致力于学习互动中的评价创新,国内集中于在线教育数据的挖掘建模。因此,应重新审视机器学习在学科教育中的应用:把准航向,开发精准技术与诠释教育本质协同共进;内生驱动,关切课程与教材研究缺失并突破政策研究困局;应用主导,打造教与学全方位的计算教育研究系统。

Machine learning has great application potential in subject education.Through sorting out the literature in CNKI and Web of Science databases in recent five years,an analysis framework is constructed to compare the similarities and differences between domestic and foreign application models:Foreign countries are skilled in technology pursuit,while domestic countries focus on implementing the original intention of teaching services;foreign countries attach importance to the management of students inside and outside the classroom,while domestic countries pay attention to the formation of learning habits;foreign countries focus on the internal emotion mining of thinking motivation,while domestic countries focus on the external emotion extraction of expression response;foreign countries advocate media coverage in and out of class counseling management,and domestic promotion of large-scale online education applications;foreign countries focus on evaluation innovation in learning interaction,while domestic countries focus on mining and modeling of online education data.Therefore,we should re-examine the application of machine learning in subject education:make progress together with the development of accurate technology and the interpretation of the nature of education;Internally driven,concerned about the lack of curriculum and textbook research and breaking through the dilemma of policy research;application oriented,building a comprehensive computing education research system for teaching and learning.