近年來運用資料探勘技術(data mining)來建構學生學習概念地圖(concept map)的方式,經常被利用來了解學生的學習狀況。在國小階段的評量幾乎都是由教師自行命題或是以教科書出版商所附之題庫光碟來編製試題,經由測驗後得到的成績來評定學生的學習成就。然而,以這樣的方式所產生的試題是否能真實的反應出學生概念學習的情況 ? 本研究提出利用一種改良的模糊關聯法則(fuzzy association rule)探勘方法,根據學習者測驗結果的數據,找出重要的關聯法則,形成較正確的學習概念地圖。接著比較藉由此種方法所產生的概念圖與專家概念圖之間的差異,來了解學生的學習狀況,同時探討教師命題的嚴謹性對建構學生學習概念地圖的影響。
Recently, many researchers use the skills of data mining to build up the concept map for students, these methods usually are applied to figure out the students’ learning condition. The method of proposition is by elementary teachers who construct questions or by textbook publishers who provide exercises disks. After the testing, that grades will become the way to evaluate students’learning achievement. However, is that a pragmatic method which reaction students’concept of learning by using those kind of testing questions? In this thesis, a fuzzy association rule method has been applied for obtaining the meaningful result of testing statistics for learners. This research can find out important relation rulers to form correct learning concept map.So, the difference between above method and expert’concept map can be compared to understand students’learning condition. Moreover, based on the proposed approach in this thesis, the completeness of the proposition can be evaluated and revealed.