摘要
近年来,人工智能和网络通讯等技术的发展使高效分析和理解教育教学过程中涌现的海量数据成为可能。由此,研究数据驱动机器学习模式下有效的计算模型和手段,从数据中洞悉教育教学中客观规律和模式,成为教育研究的热点之一。为此,文章首先回顾了教育研究所属范畴的历史变化过程;然后,文章围绕教育作为自然科学研究对象所具有的以实证经验为基础、以量化计算为手段的特点,介绍了教育过程复杂性建模、教育效果由果溯因评测以及教育实验随机对照分析三个问题中的若干计算手段;最后,文章根据现有人工智能模型解释性不强这一局限性难以更好促进教育研究进展现状,指出数据和知识双轮驱动、闭环反馈回路、随机对照实验前提假设等是教育研究技术手段今后发展的趋势,以期建立解释性更强教育研究理论和方法。
In recent years, the development of technologies such as artificial intelligence and communication network has made it possible to efficiently analyze and understand the massive data emerging in teaching and learning process.Therefore, studying the efficient computing models and means under the data-driven machine learning model and further gaining insight into the objective laws and patterns of education and teaching from data has become one of the hotspots in educational research. Therefore, this paper firstly reviewed the the historical change process of the category that education research belonged to. Further, focusing on the characteristics of taking on empirical experience as basis and quantitative calculation as the means belonging to education as a natural science research object, this paper introduced several computing methods in three problems of complexity modeling of educational process, causal inference evaluation in educational effect assessment and educational randomized controlled analysis. Finally,according to the limitation that the existing artificial intelligence models were not very explanatory, it was pointed out in this paper that two-wheel drive of data and knowledge, closed-loop feedback, and presupposition of randomized controlled trails were the trends of educational research in future, so as to establish educational research theories and methods with more explanatory.
作者
朱雨萌
钭兰朵
许馨宸
吴飞
ZHU Yu-meng;DOU Lan-duo;XU Xin-chen;WU Fei(Zhejiang University,College of Education,Hangzhou,Zhejiang,China 310058;Zhejiang University,Chu Kochen Honors College,Hangzhou,Zhejiang,China 310058;Zhejiang University of Technology,College of Computer Science and Technology,Hangzhou,Zhejiang,China 310014;Zhejiang University,College of Computer Science and Technology,Hangzhou,Zhejiang,China 310058)
出处
《现代教育技术》
2023年第2期33-42,共10页
Modern Educational Technology
基金
国家自然科学基金重点项目“面向在线教育的群体智能支持下人机协同学习研究”(项目编号:62037001)的阶段性研究成果。
关键词
人工智能
数据驱动
反馈回路
可解释
artificial intelligence
data-driven
feedback loop
explanatory
