多型態影片代理表徵對摘要影片式學習之影響; Effects of Multimodal Surrogates in a Video Summarization Learning Platform

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Network Learning Technology > Electronic Thesis & Dissertation > Item 987654321/48577

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/48577

Title:	多型態影片代理表徵對摘要影片式學習之影響;Effects of Multimodal Surrogates in a Video Summarization Learning Platform
Authors:	張文璇;Wen-hsuan Chang
Contributors:	網路學習科技研究所
Keywords:	系統使用行為;理解力;認知風格;摘要影片式學習;影片代理表徵;Comprehension;Navigation;Cognitive Style;Video-Based Learning;Video Surrogate;Video Summarization
Date:	2011-07-26
Issue Date:	2012-01-05 14:58:12 (UTC+8)
Abstract:	在過去的影片式學習環境中，學習者必須花費較長的時間觀看整部影片，或者使用影音播放器中簡易的快轉、倒轉、暫停方式來捕捉影片中的重要內容，本研究欲透過影片摘要技術來協助降低影片式學習中大量的時間成本。因此，以摘要式影片技術結合影片代理表徵(Video Surrogate)之概念作為基礎，將保留重要內容的影片摘要搭配具有輔助內容理解的訊息元件，來發展一套摘要影片式學習之平台，希望能夠藉此提供充分的學習資訊與支援工具，輔助學習者對影片內容進行有效地理解，並討論摘要影片式學習之效果。本研究透過不同的影片代理表徵組合型態來輔助學習者影片內容理解之實驗，主要探討所設計的快進播放型、影片摘要型、關鍵字影片摘要型等三種影片代理表徵組合型態，對於輔助內容理解之學習成效與差異，並且討論學習者的個體差異與系統使用行為之間的關係，以及這些變項對於摘要影片式學習理解成效之影響。最後，評估學習者知覺該系統平台的滿意度與建議，並整理學習者使用多媒體的背景因素與其滿意度間的關係。實驗結果證實，學習者使用此三種影片代理表徵後皆有達到理解之目的，而其中以關鍵字影片摘要型與影片摘要型影片代理表徵組合型態較快進播放型之理解效果為佳。此外，結果發現學習者的認知風格會影響其系統使用行為，然而並無直接影響理解成效；然而，從學習者之系統使用行為發現，學習者經由使用平台中特定訊息元件之行為，對於影片內容的理解成效有正面的效益。在系統滿意度問卷的分析結果中，學習者知覺最為滿意之組合型態為排序單型影片代理表徵組合型態，不過同時也認為關鍵字影片摘要型組合型態在有限時間下為有效抓取重點之學習工具；除此之外，經分析後發現學習者的多媒體偏好與經驗與其知覺系統有用性、易用性或滿意度之間有著正向的關係。 The traditional video-based learning seldom considers the issue of time constraint that learners may keep viewing videos until the whole video stream finished or the suitable videos are explored. They spend a lot of time grasping the important gist of videos, but yet hardly attain the learning goal. In order to learn by video summaries, this study combines the idea of video surrogate with video summarization technique as a video-based learning platform to enhance learners’ comprehension. Consequently, three composed sets of multimodal video surrogates, namely the fast forward (FF), the video summarization inventory (VSUM) and the keyword-based video summarization (KVSUM), are proposed to investigate the effects on learners’ comprehension towards video contents. Furthermore, this study attempts to explore the relationships of cognitive style and navigation in this platform, and the effects on comprehension performance. Finally, a satisfaction questionnaire is implemented to assess learners’ perceptions of multimodal video surrogates. The results reveal three possible implications: First, the video summarization learning platform with multimodal video surrogates provides helpful information to enhance learners’ comprehension. The effects of KVSUM and VSUM are better than FF. Second, the learners’ cognitive styles affect their navigation behaviors and strategies, but yet no significant effect on comprehension performance. Moreover, among those usage behaviors, learners who have utilized the specific surrogates in the platform show positive correlations with better comprehension to video contents. Last but not least, in terms of system usage and satisfaction, VSUM is most attractive to learners, but meanwhile learners think the KVSUM may be a useful tool for quick grasping gist in a limited period. Furthermore, the learners’ preferences and prior experiences of multimedia show positive correlations with usefulness, ease of use and satisfaction of the multimodal video surrogates in the video summarization learning platform.
Appears in Collections:	[Graduate Institute of Network Learning Technology] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	787	View/Open

社群 sharing

Loading...