應用資料探勘技術探討高等教育申請入學條件之研究

Using Data Mining Techniques in the Study of Student Admission Criteria in Tertiary Education

高等教育發展於專業人才培育中是不可或缺的一環，且需能符合聯合國教育科學文化組織（United Nations Educational Scientific and Culture Organization, UNESCO）所提出的公平（equity）、適切（relevance）、卓越（excellence）政策來培育專業人才，故高等教育機構所自行制定之入學標準，應建立適切的選才標準來公平地審核入學資格，並使得錄取之學生經過各系所適切的課程訓練後能有專業卓越的表現。因此，本研究旨在探討入學標準是否符合系所特色，且欲瞭解學生特質對於系所課程表現的影響，進而瞭解學生潛質是否符合該系所之特色，以達成適性揚才之目標。本研究利用資料探勘中分類（classification）、分群（clustering）、關聯（association rules）及屬性選擇（attribute selection）之技術，發掘影響學業表現之因子，並歸納與建立規則模型，進而依據此模型提供招生及決策人員客觀之建議。

Higher education is critical to professional development, and must follow the policy, according to the United Nations Educational, Scientific and Cultural Organization, of fairness, appropriateness and excellence in the development. In this regard, universities and departments should select students on a fair basis with appropriate selection criteria in order to recruit the right students to become excellent in the specific professional domain. Therefore, this study focuses on the admission criteria to see how it aligns with the department educational feature for recruiting the appropriate students. Specifically, this study uses classification, grouping and attribute selection in data mining to discover potential influential factors of applicants that have positively association with the performance of future study. Further, a model is proposed in this regard to provide objective and insightful suggestion for the consideration of admission.