電機資訊學院: 電機工程學研究所指導教授: 王凡吳啟允Wu, Chi-YunChi-YunWu2017-03-062018-07-062017-03-062018-07-062016http://ntur.lib.ntu.edu.tw//handle/246246/276438在網路普及的時代，各式各樣用途的網頁相繼開發出來，人們的生活也越來依靠各種網站。因此對開發者而言，為了確保網站程式的品質，如何快速有效的進行網頁測試就顯得重要。我們在此篇論文中開發了一套工具來自動化測試網頁，使用者不需撰寫測試腳本，就可以點擊和填值的方式自動瀏覽動態網頁，並且記錄下瀏覽的紀錄。然後我們提出了一個方法來建構測試準則，我們應用了機器學習中支撐向量機的技術，從網頁和手機的使用軌跡中抓取特徵值，順練出預測模型來自動判斷測試紀錄是否通過。藉由這個工具和驗證方法，我們可以有效地降低人力成本達到自動化測試的目的。In the recent days, the Intertnet is becoming more popular. A wide range of web applications have been developed. People spent lots of time on Internet. Thus, it becomes an important problem to verify the web applications to developers. In this paper, we propose a tool for automated web testing. The developers do not need to write the test scripts. The tool can explore the dynamic webpages by clicking buttons and insert values and record the test traces. We propose a system to construct test oracle of web applications and mobile applications using the support vector machines. The system extracts features from the traces and builds a redictive model to classify the passed traces and failed traces. With the automated testing tool and system, we can effectively reduce human cost to achieve the purpose of automated testing.1933611 bytesapplication/pdf論文公開時間: 2016/8/24論文使用權限: 同意有償授權(權利金給回饋學校)軟體測試網頁測試自動化測試測試準則Software testingWeb testingAutomated testingTest oraclethesis10.6342/NTU201601810http://ntur.lib.ntu.edu.tw/bitstream/246246/276438/1/ntu-105-R03921032-1.pdf