網際網路的普及率日益成長，線上客戶服務及網路化教育訓練的發展也愈來愈受到重視。由於目前網路化教育訓練系統的主要功能在於提供教學內容，對於學生在學習過程中遭遇的困難，仍有賴教師以人工方式個別處理：而選修線上課程的學生往往人數眾多，不僅造成教師在解答問題方面的負擔，更可能因長時間等待解答而影響學生的學習成效。針對這個問題，本論文提出一套具自我調適功能之線上課程問題自動回覆系統，可接受學生線上發間，並自動回覆正確的解答。同時，系統在回覆學生的問題時，會依據學生反映滿意程度來調整問題分析機制;當資料庫中不存在該問題的解答或學生對該問題的解答持續表達不滿意時，系統才尋求教師的協助，並將教師提供的問題與解答納入資料庫。經實驗結果發現，自我調適機制可使系統回覆問題的正確性提高。因此，本系統的成果不僅可大幅減少教師解答問題的負擔，並可24小時提供即時解答服務，減少學生在學習上的阻礙，達到增進網路化教育訓練成效的目標。

The popularity of the Internet has facilitated the derivation of information, which encourages the development of on-line learning systems. Nevertheless, most of the existing e-learning systems are not capable of assisting students while they encounter problems during the learning process; instead, they only simply provide subject materials for browsing. Therefore, students are likely to be stuck by some problems, and hence their learning performances could be significantly affected. For a popular on-line course with thousands of stu­dents, it is almost impossible for the teachers or the teaching assistants to answer all the students' questions manually, which is not only inefficient, but also human recourse-consuming. To cope with this problem, in this paper, we present a self-adaptive virtual tutoring assistant system, which is able to answer students' questions automatically based on the answers given by the teacher. Moreover, the system is able to accept the students' feedbacks and adjust the weights of keywords for each candidate answer, and hence more accurate answers can be provided in the future. Experimental results have shown that, the system can provide more accurate answers by employing the self-adjusting approach. Moreover, as such systems offer student question-answering service 24 hours a day, learning performance of the students can be improved by reducing their discouragements during the learning process in time.