【摘要】 针对Web测试中现有Web页面聚类方法无法准确描述复杂页面结构、页面聚类准确度低、时间复杂度高的问题,分析了Web页面的DOM结构和节点属性,给出改进的树匹配算法衡量Web页面间相似度,并提出一种新的标记聚类方法实现Web页面聚类。通过实验对比验证所提出的方法能够有效处理复杂Web页面结构,且聚类准确度高,时间复杂度低,是一种高质量的Web页面聚类方法。更多还原

【Abstract】 Page clustering is an extremely effective method to reduce the number of redundant state in Web applications through clustering similar pages. Page clustering needs to analyze the similarity between two web pages,but traditional clustering methods can not accurately describe the complex page structure,high time complexity and low clustering accuracy. Therefore,this paper proposes an improved tree matching algorithm,which considers not only the structure of DOM tree,but also some attribute information of DOM,which makes this method better cope with complex web page structures. Experiments show that this paper proposes an effective page clustering method,which greatly reduces the clustering time and improve the accuracy.更多还原