國高中生閱讀理解網際網路提供的各類圖文及相關問題解決: 學習動機、線上-離線認知歷程及注視驅動鷹架-子計畫三:眼動資料分析系統開發與基於眼動之人機互動研究

Abstract

利用眼球追蹤儀分析與探討人如何利用眼睛對周遭事物進行探索跟學習，在各領域 中變得越來越常見，如何對眼動資料進行正確且有效的分析變成研究的關鍵。但市面上 眼動資料分析系統的軟體和硬體部分通常成套販售且價格昂貴，對眼動研究及應用的普 及性造成阻礙。近年來許多研究開始嘗試利用價格低廉的攝影機開發低成本眼球追蹤 儀，但在軟體方面缺乏功能完整的分析軟體。而隨著網路、個人電腦和行動裝置普及化， 數位學習提供更多的彈性，讓人們可以不論何時何地，藉由電子裝置進行學習，因而受 到越來越多人重視。

本計晝將規劃眼動資料分析與基於眼動之人機互動研究平台，研發眼動資料分析系

統讓研究者可以作完整眼動資料分析並獲得研究所需之各項指標。將前述之系統進一步 應用在人機互動的研究和系統開發。本計晝將研究學習者在進行學習時的眼動模式與專 注程度之間的關係，設計專注程度偵測演算法，搭配眼球追蹤儀即時監測學習者的凝視 位置等資料，了解學習者的學習情況並由系統主動給予適當的提醒，藉此提升進行學習 者的學習成就。有了這些基於眼動之人機互動的探索與經驗後，此外有鑑於目前數位學 習對身障生的支援度低落，本計晝於第三年將著手開發適用於身障生的眼動註記系統。

The goal of the three year project aims to design, build, and implement a complete Eye movement data analyses and interaction system. This system will integrate several tools in order to achieve its goal such as the annotation tool and a modified version of the open source software TAUPE. This system will be used to test a group of students (7th grade - 12th grade) in which an eye tracker will record the eye movement of the students, and the system will analyze the recorded data.

The project is motivated in the first year to study the way the students read the learning material provided to them, and observe the way that each student follow to catch the important ideas. In this year will focus on static AOI functions. More specifically, different levels of AOI will be defined based on the importance of the learning materials to understand and observe the relationship between the eye movement behavior of students and learning materials in key areas. In addition, we are going to design "degree of dedication detection algorithm, using eye tracker to collect eye movement when students learn to analyze and determine the focus of the students in the learning process. In the second year, we will develop basic dynamic AOI definition tools, reading focus reminder system, and gaze scaling interactive system. In the third year, advanced dynamic AOI definition tool will be developed. Moreover, a tool to help the disable students will be developed.

To achieve this goal our system will integrate several tools together; the annotation tool will be used to define the AOI. Therefore, we will be able to check whether the important part of the learning contexts was read or not. This data also will be used to perform the performance analyses (analyze the gaze time percentage of each part including important and non-important parts), the taupe software will be used to visualize and analyze the recorded data. In addition new tools will be developed to achieve our goals.

We hope that our experimental results will give students a new viewpoint about learning materials, and boost new social services based on those results.