Research And Application Of Business Intelligence Technology Based On Multidimensional Dataset

As more and more fierce competitions emerge in the business world, the capability of processing and utilizing information is the key factor to enterprise's survival. Making full use of and mining the existing data in a company can help decision-makers make more agile and reasonable business decisions. So more and more enterprises and governmental organizations have applied BI products and its services to process the most important asset-information.During the design and implement for BI solution, we must review the actual conditions of enterprises' data-analysis application firstly, in this way, the deployment of BI solution can receive the expected outcome. However, the existing BI solutions have some shortcomings in their function, architecture, and applied mode. They can't meet the different needs of different enterprises and even the different information consumers well. They are also incapable of providing an excellent expansibility for the changing of BI applied level.Because of the shortcoming in existing solutions, the paper puts forward an improved BI application architecture which makes the combination of analysis services and report services come true. The architecture can support customized implement and distributed deployment according to analysis applied level of enterprises, and provides enterprises with inside expansibility to append new analysis services. The paper introduces the logical composition and feature-sets definition of the new architecture, meanwhile, it also researches the realization of the partial key-functions and analyzes the applied results of the new architecture. The content is as below:1) Based on BI knowledge, the paper discusses the actual conditions of data-analysis application in enterprises, and then it points out the disadvantages of existing BI solutions according to the diversity and variability of enterprises' requirement-level in the actual application.2) The paper introduces a new BI applied-architecture which integrates multidimensional data-analysis services and report-sharing services. Based on the chief design-thread of the new architecture, the paper describes the design of the feature sets and logical structures of the architecture's components such as multidimensional-analysis platform, analysis-report server and so on.3) According to the character of users' analysis behavior and the restriction of incremental operation based on the non-structure of Multidimensional Expressions, the paper puts forward an object-oriented query structure. Meanwhile, based on the structure, the paper also expounds the solution of tracing analysis clues and serializing analysis results.4) Combining ADOMD.NET data-provider with XML for Analysis specification, the paper does some researches on how to access and manipulate data and metadata of multidimensional data sources in .NET platform. Meanwhile, it sets forth the design and realization of the function for multidimensional data access in the improved solution.5) Based on the design-requirements of report-sharing services and the advantage of developing Web Service application in .NET platform, the paper sets forth the realization of issuance and browse of analysis report basing on Web Service technology in the improved BI solution.