随着移动应用市场的快速发展，移动应用兼容性测试问题日显突出和紧迫。本文提出了一种基于K⁃Mena算法的移动应用兼容性测试设备选择方法。该方法首先建立移动应用兼容性测试设备的特征树模型，确定各个基本特征的测试值，定义初始K值和中心点。然后给出一种基于设备特征树模型的特征距离计算方法。最后基于K⁃Mean聚类算法，对移动设备进行聚类，从各个聚类中选择适合的移动应用兼容性测试设备。通过实例进行验证分析，结果表明该方法能够帮助测试人员选择适合的测试设备，从而降低测试成本，提高测试效率和测试质量。%With the rapid development of mobile application′s market, mobile application compatibility testing has become a most prominent and urgent problem. We propose a systematic method of selecting adaptive mobile devices for mobile application compatibility testing;it is based on K⁃mean algorithm. Firstly, the compatibility feature tree model of mobile application is constructed, and every leaf node of feature tree is set some testing values by experi⁃ence. So initial value of K and cluster centers are set. Secondly, a layered and recursive algorithm is described for calculating feature distance between mobile devices. Thirdly, all mobile devices can be clustered automatically with K⁃mean algorithm, and testing devices can be selected from different clusters according to popularity market share of devices. Finally we test two selected mobile applications with proposed method, proving that the method is effec⁃tive in selecting mobile devices for compatibility testing of mobile applications.
展开▼
