資料探勘於課程選擇之應用

Data Mining on Course Selection

楊裕祥(Yu-Hsiang Yang)

指導教授 ： 吳建文

國立臺北科技大學/管理學院/工業工程與管理系所/碩士(2005年)

摘要

在現實環境多重限制之下，如何適當且方便選擇出，使大多數學生滿意的課程集也就是滿意課程集，是一個非常複雜的課程選擇問題，而此問題確是存在於每一所學校當中。
目前在解決課程選擇問題上，皆是使用機率為基礎方法，來決定所要設立課程集，但這並未考慮到其選擇的課程集，是否能滿足學生的需求，換言之以目前開課方法，並不能保證所選擇出的課程集，能得到大部份學生的滿意，也就不一定是滿意課程集。
故本研究嘗試運用文獻上已有的滿意項目集探勘(Satisfying Itemset Mining)來進行滿意課程集探勘。本研究以國立台北科技大學工業工程與管理系在職班，兩個年級為實驗資料，並與目前機率為基礎的選課方法進行比較。實驗結果顯示，本研究所運用的方法，進行課程選擇後所選出的課程集，在學生滿意度及課程彈性方面，都比現行機率基礎的決定課程集方法皆有較佳的表現，證實本研究方法所選擇之課程集，是可以滿足大多數學生需要的滿意課程集。在此可提供學校在未來課程選擇方法上一個新的選擇。

關鍵字

課程選擇問題 ； 資料探勘 ； 高頻項目集探勘 ； 滿意項目集探勘

並列摘要

Under constraints of the realistic environment, how to select a course set that satisfies a large percentage of the students is a very complicated course selection problem faced by every school.
The current approach uses probability as a base method which can not be guaranteed to find satisfactory results for complex problems. Nor it considers whether the course set satisfies students’ demand. In other words, the current approach provides very limited solutions to satisfy students’ demand in the course selection problems.
This study applies satisfying itemset mining that existed in documentation to discover a satisfying course set. The study uses the part-time students’ data to compare the satisfaction and flexibility between the approach adopted by this study and the probability based approach. The research results reveal that the proposed approach within this study would be able to generate a more satisfactory and flexible satisfying course set than the probability based approach.

並列關鍵字

Course Selection Problem ； Data mining ； Frequent Itemset Mining ； Satisfying Itemset Mining

參考文獻

[1]. Agrawal, R., Imielinski, T., and Swami, A. 1993. Mining Association Rules between Sets of Items in Large Databases. Proc. ACM-SIGMOD Int. Conf. Management of Data (SIGMOD’93), Washington, USA, pp.207-216.

[2]. Agrawal, R., & Srikant, R. 1994. Fast Algorithm for Mining Association Rules. Proceedings of 20th VLDB Conference, pp.487-499.

[4]. Chienwen Wu 2004. Application of Frequent Itemset Mining to Identify a Small Subset of Items that can Satisfy a Large Percentage of Orders in a Warehouse, accepted in Computers & Operations Research.

[6]. Han, J., Pei, J., & Yin, Y. 2000. Mining frequent patterns without candidate generation, Proceedings of the ACM SIGMOD International Conference on Management of Data, Dallas, Texas, USA, pp.1-12.

[7]. Han, J., Pei, J. 2000. Can we push more constraints into frequent pattern mining?, Proc. Int. Conf. Knowledge Discovery and Data Mining (KDD’00), Boston, MA, pp.350-354.