網路新聞的蓬勃發展,使得民眾得以輕易且迅速的取得新聞資訊,但卻因資訊量過於龐大、難以消化,進而衍生資訊的可搜尋性問題。過去的研究指出當使用者不了解其所欲尋找的資訊時,搜尋引擎無法幫助使用者找到正確的資訊,例如下關鍵字的問題。故過去亦有研究者嘗試利用自動化摘要、相關性回饋、資訊視覺化等方法試圖幫助使用者能夠獲取有效且受用的資訊。著眼於此,本研究利用動態網絡視覺化方法自行開發一套新聞互動導覽系統,系統包含利用網絡圖進行關鍵字推薦、瀏覽歷程紀錄與控制、繪圖原則切換、資訊過濾與資訊分群等功能,以引導使用者搜尋與瀏覽新聞資訊。在系統評估方面,本研究透過問卷、訪談、歷程序列分析等方法可發現本系統獲得以下正面回饋:1) 本系統利用網絡圖進行關鍵字推薦與群集幫助使用者有效挖掘相關資訊;2) 網絡繪圖規則切換能夠幫助使用者以更多不同的角度瀏覽資訊;3) 關鍵字共現強度濾器,能夠幫助使用者過濾複雜的資訊,以助於進行下一步驟之瀏覽;4) 歷程紀錄與控制功能,能夠導航使用者的瀏覽過程,並自我定位瀏覽路徑。然而,亦有使用者表示在操作性或認知上的負面回饋,例如,有使用者提及雖然系統提供相當多元的功能幫助進行新聞瀏覽,但使用門檻高,使用時不易理解種多功能所能帶來之助益。
The raising of online news allows people to obtain and access the news information easily, however, also increases the difficulty to search browse, and digest daily news. To assist users in news browsing and searching, this study built up a dynamic network visualization tool composed of visualized concept maps, trends of keywords, and interactive browsing paths, and related news contents. Users can drag and click on terms to expand related ones, traverse or alter the browsing path, and explore the concept network from different points of view. Three layout methods of visualizing concept maps were implemented to list and cluster concepts for different purposes. A user study was conducted to evaluate the usefulness and usability of our tool. The results of the user study showed both positive and negative opinions from our participants. They agreed that network-based visualizer provides related terms for the next browsing, and the interactive browsing path can help them control the set of related news. However, some of the participants indicated that the visualized features were too complex and even confusing for use.