利用關聯規則發掘學生適性課程

本研究探討資料探勘（data mining）技術協助學生選修課程的傾向性，從過去大量修課資料中發掘不同課程之間關聯性，對發掘學生適性課程的推薦，以達到學生選修適性課程、及選修課程的學生都是適合的學習目標。文中以學生的修課資料為探勘資料來源，每一筆修課資料包含學生每學期曾經修課的課程項目、及課程的開課學期，並以k位學生為探勘目標，k≧1，利用資料探勘技術中的關聯規則（association rule）發掘學生適性課程。文中設計一個方法探勘關聯規則，其前置項目集必須為k位學生已曾修讀過的課程項目，後置項目集必須為k位學生這學期可選修的課程項目，藉由探勘以上形式的關聯關規則，可分別做為發掘k位學生適性課程的依據。本研究修改Apriori演算法，直接組合k位學生的修課資料中項目與這學期可修讀的課程項目形成項目集，並判斷這些項目集是否為高頻項目集，將可更有效率找到目標項目集。本研究根據提出的方法，設計與建置一個發掘學生適性課程探勘系統，其探勘結果，對學校課程單位規畫學生個人適性課程，以主動服務方式推薦學生選修課程科目，可以提供非常有用的參考資訊。

This study explores the tendency of using data mining techniques to assist students in taking elective courses. We discover the relevance of different courses from a large number of course taking data in the past, and to recommend for finding students' adaptive courses. In order to achieve learning goals for students to take adaptive courses and students of taking courses are suitable. The paper uses the students' course taking data as the source of mining data, and each course taking data contains the course items that the student has taken each semester and the starting semester of the course. Taking k students as the mining target, k≧1, we use the association rules in data mining techniques to find students' adaptive courses. A method is designed to mine the association rules whose antecedents must be courses that k students have taken, and consequent must be courses that k students can take this semester. By mining the above association rules, they can be used as the basis for finding k students’ adaptive courses. We modified the Apriori algorithm to directly combine the items in the course taking data of k students with the course items that can be taken this semester to form itemsets, and to judge whether these itemsets are frequent itemsets. The target itemsets can be found more efficiently. According to the proposed method, a mining system for finding students’ adaptive courses is designed and built. The results of mining can provide useful reference information for school curriculum unit to plan students’ personal adaptive courses, and to recommend students to take courses in a proactive manner.