安卓應用程式自動化品質測試

Abstract

本研究介紹一系列基於App Quality Alliance(AQuA)組織所制定的應用程式品質標準，應用於安卓系統應用程式的品質自動化黑箱測試演算法，並且實作於一套測試工具App Quality Tool(AQT)之上。藉由PC端自動操縱測試裝置，模擬使用者使用應用程式的情境，並同時收集測試裝置上的相關資料，從而判定安卓應用程式是否符合AQuA組織所制定的品質標準，並且產出一份測試報告簡述測試環境與測試過程。 為了應證此自動化演算法是否能揭露應用程式之缺陷，本研究將此演算法實際運用於真實世界之熱門應用程式。實驗結果證實，此演算法可偵測應用程式在品質設計上之缺陷。透過這一系列自動化演算法，可以降低大量人力成本於測試安卓應用程式，增進應用程式開發之品質。

In this work, we investigate how to automate the test criteria specified by Application Quality Alliance (AQuA) to help to contain the testing cost for the quality of Android mobile applications (Apps). We proposed and implemented several black-box testing algorithms for several AQuA Tests on Application Quality tool (AQT), a testing tool can manipulate the testing device and collect data from the testing device. We check the data collected from testing device and carried out experiments to see how our techniques work for several real-world public mobile Apps. According to our experiment results, our algorithms can find out some faults from the real-world public mobile Apps. Finally, our testing procedure generates a testing quality report describes the result of the test for the tester to review the information and result of this test.