A Pattern-Oriented Impact Analysis Approach for Mashups
A Pattern-Oriented Impact Analysis Approach for Mashups
复制标题
DOI：
10.1109/sose.2010.17
复制
发表时间：
2010-06
期刊：
影响因子：
--
通讯作者：
Weilong Ding;Guiling Wang;Yanbo Han;Jing Wang
中科院分区：
文献类型：
--
作者：
Weilong Ding;Guiling Wang;Yanbo Han;Jing Wang
Mashup因其方便的构建和共享时尚而在最终用户中广受欢迎。不可避免的是，Mashup很容易因为其集成组件的更改而崩溃，比如升级和不可访问。本文主要研究Mashup的变更影响分析。我们从面向模式的角度提取Mashup的数据依赖关系，并将Mashup的创建抽象为依赖关系图。通过我们定义的图和影响概率，可以定量地计算组件对混搭的影响。该方法在我们的混搭环境MashRoom中实现。文中还讨论了案例分析。
Mashups have gained popularity among the end users for its convenient building and sharing fashions. Inevitably, a mashup is vulnerable to break down caused by its integrated components’ change, such as upgrade and inaccessibility. This paper focuses on the change impact analysis for mashups. We distill the data dependencies of mashups from pattern-oriented perspective, and abstract the mashup creation to a dependency graph. With the graph and the influence probability we defined, the impact of the components on the mashup can be calculated quantitatively. The approach is implemented in our mashup environment –Mashroom. The case study is also discussed.