教育性超媒體學習型態之研究

Abstract

教學超媒體系統提供了不同於傳統的知識傳授模式與方法，雖然使得教學 有效率，且更有彈性，可是其中也衍生了不少的問題。理想的超媒體電腦 輔助學習系統，除了教材本身豐富的內容外，系統本身要能分析教材網路 的優劣、並提供有助於學生建構知識的工具，及輔助學生在超媒體系統內 瀏覽。在超媒體的教學環境中，「教材的結構」與「學生的閱覽行為」是 兩個主要的研究主題。本篇論文的目的就是針對這兩個主題提供系統化的 分析工具，以便能達到理想的超媒體電腦輔助學習系統。在課程結構的主 題上，我們應用四種圖學上的演算法：最小路徑分割、強連通、關蜃I分 割、動態網路分析。透過這些演算法，可將雜亂的教材網路建立為階層式 網路，並分析教材網路的優劣，以幫助學生建構知識，並期能解決超媒體 迷失的問題。對於學生的閱覽行為，我們提出兩種量化的分析指標：延伸 型超越度(ExtensionalHyper Degree) 與超越距離(Hyper Distance)。經 由計算，我們可以客觀的了解學習者的學習模式。 Educational hypermedia provides flexible access to instructionalinformation and supports convenient learning styles.However, what it provides is a non-sequential information presentationwhich is different from traditional instructional systems.Thus, there are many problems resulted from educational hypermedia systems.In an ideal educational hypermedia system, in addition to supportingabundant courseware, it should provide courseware analytical tools,knowledge construction assistance tools, and navigation guidance tools.Furthermore, in an educational hypermedia environment, courseware structure and learning behavior are the main research topics.In this thesis, we provide several systematic analytical models in order to approach an ideal educational hypermedia system.To identify courseware structures, we adopt four graph algorithms for educational hypermedia analysis: MinimumCut-Set, Strongly Connected Components, Cut Vertex, and Dynamic Analysis.By using them, we are able to construct a hierarchical knowledge space,to analyze a courseware network to see whether it is well- structured, and to design a hierarchical map to help users learning in a hypermedia environment.To understand learning behavior, we provide two quantitative measures:Extensional Hyper Degree and Hyper Distance.According to the computational results, we are able to better understand students' navigation activities in the educational environment.