Flexible-Matching Data-Comparison Tool with the Cloud Applications

Abstract

Cloud computing provides an infrastructure that enables rapid delivery of computing resources in a scalable virtualized manner. Software residing on the clouds, often called Software-a-a-Service (SaaS), is exposed to end users as web applications in browsers or apps on thin clients such as mobile devices and netbooks. Data on clouds are often huge in size, and they are processed in the MapReduce manner in real time. A data differencer is needed to compare the data for regression testing and data filtering. However, traditional data comparison techniques for relational or semi-structured data such as XML files do not take unique issues brought by cloud computing into account. This paper presents a fine-grained flexible-matching data comparison algorithm that can compare encoded data according to users' specific interests and visualize differences in a user friendly manner. This paper uses a weather forecasting problem to illustrate the algorithm by running it on a cloud environment using Protocol Buffer data format. The experiment results show that the proposed algorithm is effective and efficient in identifying data differences in real time on the GAE.