目前大多學生課後學習以傳統習作方式學習，多為教師指定回家作業，由於缺乏互動性，以致於學生學習意願低落以及缺乏思索，等待教師告訴下一步驟，並且學生在遇到新問題時，缺乏詮釋題目的能力，而無法有效地將所學知識應用於新的問題上。 本論文建置遊戲學習系統提升學生學習動機，再以問題導向學習設計遊戲題目，透過情境式題目使學生詮釋題意後解決問題，並且學生可以檢視自我學習歷程，瞭解自身學習情形，訂立學習目標，達到自我導向學習效果。 應用資料探勘中的決策樹演算法分析學生學習表現與遊戲表現之間關聯性，透過指定因子分類模組找出影響學生學習成效因素，並提供教師作為教學建議與學生學習建議。 At present, most students do homework that is the way of traditional. Teachers design homework in most cases. It lacks of interactivity, so that students with poor willingness learn and lack of thought. They wait for teacher to tell the next step. When students encounter new problems, because they lack the ability of interpreting the subject, they can’t effectively applied the knowledge on new problems. In this thesis, using game learning system to improve learning motivation, then using problem-based learning to design game problem. Through situational problem, students can interpret the meaning of problems and then solve the problem. In addition, students can view the self-learning process and build learning goal. Finally, they are able to reach self-directed learning. In this thesis, using item response theory to analysis the response of patients in questionnaires and further investigate its causes and subsequent coping. Using of the decision tree algorithm to analysis the correlation between students'' performance and game performance. Finding the factor of student learning efficiency through specify classification module. These factors can provide teachers and students with suggestions.