摘要
界面采集技术是软件自动评测、软件训练评估领域关键技术之一。对于一些规模较大且部分界面非标准化的应用软件,会出现部分界面难以识别且采集效率低的问题。提出一种基于界面特征的识别模型,运用软件行为学理论描述界面交互行为,采用加权欧式距离与句法模式识别相结合的方法对目标界面进行匹配,较好解决了目标界面识别问题。实验测试证明该模型有效、实时性好。
Interface information collection technology is the key technology in antomatic software evaluation and software training evaluation.However,for large-scale application software with the interface being partially non-standandized,some interfaces are difficult to recognize and the collection is not efficient.To solve these problems,this paper proposes a recognition model based on interface features.Using software behavior theories to descibe interface interaction and using euclidean distance and syntactic pattern recognition to match the target interface can effectively realize target interface indetification.Experimental tests show that the model is effective and real-time.
出处
《信息工程大学学报》
2012年第2期241-246,共6页
Journal of Information Engineering University
关键词
软件行为
欧式距离
句法模式识别
信息采集
software behavior
euclidean distance
syntactic pattern recognition
information collection
