本研究開發一套適用於不同聽力理解程度學習者的英語影片學習系統，希望能夠提供符合學習者聽力理解能力且充足完整的影片情境來輔助學習者英語聽力理解，並藉由SVM演算法以自動化的方式建立影片難度分級。此外，本研究設計一項實驗來評估使用字幕特徵、聲音特徵、字幕及聲音特徵，以哪種特徵方式所分類出來的影片最符合學習者聽力理解能力，以及對於輔助聽力理解成效之差異性。最後透過科技接受度和系統品質問卷調查學習者使用本系統後，對於系統的滿意度。實驗結果發現，第一，在聽力理解能力為高程度與中程度的學習者，使用字幕聲音特徵所分類出來的影片最符合其聽力理解能力。然而對聽力理解能力為低程度的學習者，使用字幕特徵所分類出來的影片，最符合其聽力理解能力。第二，不論使用字幕特徵或字幕聲音特徵所分類出來的影片，皆有助於學習者英語聽力理解。第三，學習者對於系統的滿意度，包含系統有用性、系統易用性以及系統品質上，皆給予正面的評價。The study proposed an English video learning system adapted to oriented EFL learners with various listening comprehension. The English video learning system not only suits to learners with various listening comprehension but also offers plenty of video context to improve learners’ listening comprehension. Moreover, the learning system would classify the videos into clips with different levels of difficulty automatically according to SVM algorithm. Besides, we conducted an experiment to evaluate the effects for using features based on various attributes, such as subtitle, audio, both subtitle and audio. The study validated that which feature suit and help learners most on listening comprehension, as well as the differences between them. We made use of the questionnaires of Technology Acceptance Model and System Quality to evaluate the learners’ satisfaction toward the learning system after using it. The results of the experiment are shown as follows. First, the high and medium level learners suit to the video clips classified according to features of both subtitle and audio most. However, the low level learners suit to the video clips classified according to features of subtitle most. Second, the video clips classified according to features of both subtitle and audio is contributive to facilitate learning comprehension. Third, the learners show positive on usability, easy-to-use, as well as quality of the system.
