在龐大的成績資料庫中，隱藏了許多我們未曾發現的重要資訊與知識。利用資料探勘技術，可以萃取出學業成就之間的關聯性，進而掌握低學習成就的學生群體，及其在學業表現上的特性，用以支援教師在課業輔導活動中的決策過程。在課業輔導的目標下，本研究分析了傳統的關聯規則與決策樹方法不適用的原因，並提出一個「特定項目關聯規則」的探勘演算法，來滿足此目標的需求。這個演算法允許使用者將某科目的低成就群體，設定為感興趣的特定項目，經由演算法探勘出其與其他科目在學業成就上的關聯性。我們以實際的個案為例，說明了這套方法在實務應用上的實施過程，同時驗證了它的可行性。而所產生的規則，不但可以有效支援課業輔導活動，更可以作為課程規劃上的參考依據。

There hides a great deal of important information and knowledge that we have never known in huge score database. In application of data mining technology, we may extract the connection of academic achievement, and then know well about student group with low academic achievement, and its characteristic of academic performance. The data mining technology is, furthermore, applied to support teachers' decision making in student counseling activities.With the goal of student counseling, the analysis of study provides the reason why conventional association rules and decision tree method are not proper, and also present the mining algorithm of item-specified association rules to satisfy the demand of the goal. This algorithm allows the user to set up certain subjects with the low achievement group into interesting specific items and by the help of mining algorithm, the connection of academic achievement can be successfully explored in other subjects. Take a practical example, we illustrate the process of this method in the practical application, and verify its feasibility at the same time. The rule produced not only can effectively support student counseling activity, but also can be used as the curriculum planning as the reference.