Mondrian Based Business Intellence System Design And Implementation

This paper studies the business intelligence related technologies, such as ETLtechnology, open source OLAP engine Mondrian, data mining, etc., especiallyMondrian system architecture, the core algorithm, and implementation of XMLAinterface methods, and improved the way of Mondrian front-end presentation toolJpivot.After study these relevant theories, using the J2EE specification and designpatterns to develop a B/S architecture BI portal system, namely business intelligenceOLAP data analysis section, other more, it integrated messaging, workflow, maps,information dissemination and common web application functions. System codereference Mondrian architecture and XMLA implementation. Reception show is amultidimensional reporting system, include interface design, statements and informationbe presented, process control, report management, system management, Systemworkflow is done by sending the data warehouse XMLA MDX queries,multi-dimensional data sets show and analysis. Data presentation and analysis methodsusing data tables and Flash graphics, and charts and graphs to achieve the data linkage.In addition, the system also implements the basic rights management, data interface,data import and export functions, provide a variety of data analysis tools, based onanalysis results, analysis or production operations to quickly generate reports. Theentire system development using the Eclipse IDE, front-end JavaScript framework usesdojo, background is mainly developed using the Spring Framework. The entire projectcan be deployed on Tomcat, Jboss and other servers, JDK1.6or above, variousWindows platform.Data extraction, data warehousing, data computing，data analysis and show consistall technical stack of BI system, this system implements the core function of BI, namelyOLAP part, the realization of the multi-dimensional data, data analysis and presentation,it has great application value in the market. According to the actual needs of themedical industry, analyze and abstract the corresponding business topics and analysis ofindicators. Through data extraction, data modeling, to achieve a common dimensiondrag and drop and OLAP slice, drill-down capabilities for data analysis to provide avariety of data analysis ways. Customers give good comments on the system after it applied to product environment.