The Research On Architecture Of Enterprise's Business Intelligence DSS

With the rapid development of network and information technology,enterprises produce and store large number of data including client data,sales data and finance data every day.Information processing system had already collected great amount of data, which is increasing exponentially,and the scattered data distributed in different databases formed separated data silo.Enterprises urgently need analyzing data efficiently,exactly and scientifically,mining potential chance in order to find out its meaning.Thus, Enterprises will understand their running situation and outer environment and make scientific decision and therefore win in modern fierce competition.Presently,the implementation of business intelligence still rest on departmental practice to a large extent,with the low status of business intelligence.In the early time of architecture of business intelligence system,interrelated people usually ignored BI system, which was short of money.Due to BI's alone island,architecture of BI system lacks self-contained theory support and uniform standard.Because there are different kinds of BI system in a company,and we need integrate such alone data to fit intelligent development of information management better,this paper try to set up the frameworkwork and approach of enterprise's BI decision support system.In such an enterprise's BI decision support system,all staff including outer participants will be able to gain information and share knowledge,which shows collective wisdom.Enterprise's BI decision support system can also help solve the problem of data storing when information is under such a disordered state.Enterprise BI system is a complicated project.Enterprise's BI decision support system is a new applied fields,the author had studied a lot of literatures and further summarize information on the basis of others' results.So this paper will be helpful to apply Enterprise's BI system with the academic knowledge studied before.This project based on three big information technologies:data warehouse,data mining and online analysis,in which online analysis consists of relational and multi-dimensional model.There are seven step to be taken in the system construction,including:requirement analyze,data pre-process,data warehouse model construction,decision support,commercial index display,training and communication,system maintenance and management.In the process of analyze and decision of BI implementation,the parts above combined to work together while realized the whole procedure from getting data,analysis,mining and decision support.The paper also used several methods such as:ROI evaluation,flurry evaluation,user satisfaction evaluation and performance evaluation to evaluate the implementation of architecture of Enterprise BI system.Nowadays,many enterprises have established departmental data mart,and simple front-end display system.As the growing demand of data,some enterprises will establish enterprise class BI decision support system.The paper analyzed the two cases in the end, and that could be used as theory and reference in practical application.