Client-centric mobile OLAP

As the architecture for clients of a client-server OLAP system, the client-centric system distinguishes itself from other clients in the marketplace by incorporating a light-weight data engine, such that the client is capable of answering queries offline when the related data are cached in the memory. In a previous study of client-centric OLAP system, a web-based client is built. Three different data visualization scenarios have been identified for which the client can pre-load data into the client as a sub-cube from the server, so that the user may drill-down the sub-cube, with the help of the data engine. In this research, the app-based version of the clientcentric OLAP is implemented. It runs on an android tablet featuring a multi-core CPU. From the comparative performance data, it is shown that the app-based client is superior to the web-based client in performance in most computational tasks of the three data visualization scenarios, especially when a parallel version of the app-based OLAP is deployed.