相梦彤 付 寒 田济榕 张婧婧
（北京师范大学 教育学部，北京 100875）
【摘 要】本研究聚焦可视化编程环境下新手进行NLP问题解决的探索过程，对学习行为进行分析以揭示视频
学习对NLP问题解决的支持作用。结果表明，视频学习平台中回看视频、自我报告等行为对新手问题解决具有明显
的提示和纠错作用，学习者在可视化编程环境下的问题解决面临模块参数编辑、变量应用的主要困难。研究还发
现了新手无目的地探索和反复尝试行为在学习中的有效性，如何转变学习观念，找到学习者自由探索与教学支持
干预的平衡点，是面向“做中学”的人工智能教育中设计和搭建有效脚手架的关键。
【关键词】可视化编程；视频学习；问题解决；学习行为分析
How to Use Video Learning to Improve Novices’ NLP Problem Solving in the Visual Programming
Environment
XIANG Mengtong, FU Han, TIAN Jirong and ZHANG Jingjing
(Faculty of Education, Beijing Normal University, Beijing 100875, China)
Abstract: This study examines how novices solve NLP problems in an online visual programming environment. Using learning analytics,
it is possible to see how online video tutorials could assist novices in NLP problem-solving. Behaviors associated with video learning,
like wise rewatching videos and reporting one’s own progress in an online forum, have a significant influence on finding inspiring
solutions and rectifying mistakes. When learning visual programming, learner’s most challenging attempts are modifying parameters
in blocks and manipulating variables. A key consideration in the development of educational scaffolds for artificial intelligence could be
how much leeway learners are given for unrestricted exploration while yet providing them with adequate instructional support.
Keywords: visual programming; video learning; problem solving; learning behavior analysis
下载： 
可视化编程环境下视频学习如何促进新手NLP问题解决.pdf
