- 作者: 林富雄; 荊宇泰
- 作者服務機構: 國立交通大學資訊科學系
- 中文摘要: 我們提出了一個好的方法來定義在超文件系統的相似性,這個工作的動機在於當學生使用超文件教學系統時定義他們之間學習路徑的相似性,最大相同子路徑和很多其他特性之差異被對應到一個實數以表示其相似性,我們證明了,如果超文件系統能夠被表示成一個非循環的有向圖,則最大相同子路徑能夠很快的被獲得;對於一般的例子來說有向圖是會有循環的,我們建議使用公制來定義相似性,釦制是基於最大相同子路徑及其他特性而被表示的,學生學習路徑則能夠藉由分群法中更進一步的應用來分析。
- 英文摘要: We propose a quantitative approach to determining the similarity of navigation pat-terns in a hypertext system. The motivation of this work is to define the similaritybetween the learning patterns of students when they use a hypertext tutoring system.The longest common subsequence and the differences of many other attributes are mappedto real numbers to represent the similarity. We show that if the hypertext system can berepresented as an acyclic digraph, the longest common subsequence can be obtainedefficiently. For the general case that the digraph is not acyclic, we propose to usemetrics to define the similarity. Metrics based on the longest common subsequence andother attributes are presented. Students' learning patterns can be analyzed by furtherapplication of clustering methods.
- 中文關鍵字: --
- 英文關鍵字: --