人工標記註解介面設計在第二外語閱讀輔助之效果評估

Evaluating the Effects of Interface Designs with Human-Generated Annotation Support for Second Language Reading

韓政憲(Cheng-Hsien Han)

指導教授 ： 王浩全

國立清華大學/電機資訊學院/資訊工程學系/碩士(2016年)

摘要

在資訊過剩的時代，線上閱讀內容的提供量遠遠超過讀者可以自行負荷的能力。這一類資訊過載的問題，則可能在遇到語言障礙時加劇。在探索提升閱讀效率的方法上，我們認為，註解對於在設計第二外語閱讀界面上扮演著相當重要的角色。這裡，我們探討人工生成的註解作為閱讀第二外語文章輔助的作用。在前兩個研究中，我們探討使用圖像作為輔助來視覺化第二語言及機器翻譯後文章中關鍵字做為閱讀策略對於讀者之影響。研究一的結果表明，相對於傳統的閱讀和學習策略（例如：中文定義輔助），使用與語義相關的圖片，讀者在長期能達到顯著較高的理解。在研究二中，我們發現，相較於使用原始圖片作為輔助，具有區域標記的圖片可以更進一步地輔助讀者對於機器翻譯文章的主觀理解。這一發現表明利用圖像於介面設計可以作為文章理解補充註釋的替代方法。在研究一和研究二之後，我們試圖了解讀者如何看待其他讀者產生的註解以降低讀者所需的認知負荷。故在研究三，我們引進群眾外包的技術以模擬社群註記環境來視覺化文章的重要區域，試圖提高讀者的閱讀效率。通過一系列的研究，我們提出了基於前述研究得到之設計寓意和準則的第二外語閱讀介面設計，努力往減低跨語言閱讀障礙的目標努力。

關鍵字

第二外語閱讀輔助 ； 人工註記 ； 群眾外包 ； 圖像筆記 ； 區域標註註記

並列摘要

In the era of information overabundance, the amount of content available to read and need to read may easily exceed what readers can handle on their own. The problem of information overload can be exacerbated when encountering language barriers. In exploration of methods for improving reading efficiency when dealing with second language content, we consider that annotations can play an important role in interface design. Here, we present studies that examine the effects of human-generated annotations as reading support for second language articles. In the first two studies, we explore the strategy of using images as a secondary and supplementary medium to visualize keywords within second-language articles as well as machine-translated passages. Results of study 1 indicated that with semantic-related pictures, readers reached significant higher comprehension in the long term, comparing to traditional reading and learning strategies (e.g. Chinese definitions as support). In study 2, we found that region-of-interest annotated pictures can better support readers’ subjective comprehension than using raw pictures as support. The findings suggest an alternative design of applying pictures as a supplementary resource in textual comprehension. After study 1 and study 2, we begin to investigate how readers perceived annotation support generated from other readers. In study 3, we apply crowdsourcing technique to simulate the social environment and visualize the important part of article in order to improve readers’ reading efficiency. Through a series of studies, we propose a reading system design that is developed based on the design implication and guidelines from the findings, to work towards the goal of crossing language barriers.

並列關鍵字

second language reading support ； human-generated annotations ； crowdsourcing ； picture note-taking ； region-of-interest annotations

參考文獻

Barsalou, L. W. (2008). Grounded cognition. Annu. Rev. Psychol., 59, 617-645.

Childers, T. L., Houston, M. J., & Heckler, S. E. (1985). Measurement of individual differences in visual versus verbal information processing. Journal of Consumer Research, 12(2), 125-134.

DeStefano, D., & LeFevre, J. A. (2007). Cognitive load in hypertext reading: A review. Computers in Human Behavior, 23(3), 1616-1641.

Francis, W. S., Fernandez, N. P., & Bjork, R. A. (2010). Conceptual and non-conceptual repetition priming in category exemplar generation: Evidence from bilinguals. Memory, 18(7), 787-798.

Gao, G., Wang, H. C., Cosley, D., & Fussell, S. R. (2013, April). Same translation but different experience: the effects of highlighting on machine-translated conversations. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 449-458). ACM.