标题: 裁判式学习: 建构式电脑辅助学习之研究
Learning by Judging: A Constructive Computer Assisted Learning Approach
作者: 徐正群
Cheng-Chun Hsu
孙春在
Chuen-Tsai Sun
资讯科学与工程研究所
公开日期: 1994
摘要: 主动学习的重要性已是教育工作者的共识, 如何提升学习者的主动性是近
年来许多研究的主题. 在本论文中, 我们提出一种以建构主义为基础的新
教学策略,称作[裁判式学习](Learning by Judging), 并且发展了一个电
脑辅助学习的环境来实现建构论中的几个关键概念, 例如引导知识重组强
化.意识后设认知.促进社群共识等等.此教学系统是一个合作式学习环
境, 学生可于此工作导向的使用者介面中 , 呈现他们的先备知识. 同时
我们采用遗传演算法(Genetic Algorithms)以提供演化样本空间, 让学生
可以反覆评估富有变化的样本. 最重要的是 , 整个学习过程中, 学生只
须做判断即可, 这个极为单纯的条件可有效的降低学习障碍, 促使学生主
动投入学习过程. 透过演化式的判断过程, 系 供学生建构与再建构其知
识的方法.为了分析学生的判断行为, 系统记录了学生所有学习过程中所
做的判断.而所收集到的量化资料将利用人工智慧的方法,如阶层式的分群
法( Hierarchical Clustering ).模糊分群法(Fuzzy Clustering). 及模
糊专家系统(Fuzzy Expert System) 等加以分析. 本方法的整体目的是要
鼓励学生将认知的过程, 及社群互动的影响具体化. 换句话说, 学习过程
的分析结果可以来解释知识建构过程, 及了解在一个网路化合作学习环境
中的社会学习型态.
In this thesis, we propose a new,constructivism-based
pedagogical strategy, called learning by judging, and develop a
computer assisted learning environment to realize this concept.
This tutoring system is a cooperative social learning
environment in which students present their prior knowledge on
a task-oriented user interface. The system employs Genetic
Algorithm to provide an evolutionary judgment process, the
system provides the students a way to constructing and re-
constructing their know- ledge. In order to analyze a student's
behavior of adjustment ,the system records his entire learning
process in the form of judgment actions. The collected
quantitative data will then be analyzed by using computational,
intelligence methods such as hierarchical clustering , fuzzy C-
means clustering , fuzzy set and fuzzy expert system. The
overall purpose of this approach is to encourage students to
realize own cognitive process and the impact form social
interaction. In other words, the analytical result of the
learning process becomes the basis of explaining constructive
epistemological processes and understanding types of social
learning processes in a network-based cooperative environment .
URI: http://140.113.39.130/cdrfb3/record/nctu/#NT830394017
http://hdl.handle.net/11536/59037
显示于类别:Thesis