Enterprise Data Analysis System Developed

At the end of the 20th century, with the development of the aggravation ofthe market competition and information social demand, it is more and moreimportant that we make market message of tactics from a large number of data(select, query, etc). This kind of demand requires the on-line service, also involve alarge number of data which are used in decision, but traditional database system isalready unable to satisfied with these kind of demand. The traditional database isreplaced by the data warehouse gradually, with the development in technology suchas database technology, artificial intelligence and mathematical statistics, DM (DataMining) technology prevails day by day, data warehouse foundation of platformoffer the good operating environment for DM technology of analyzing in data, makeit to play more and more greater research potentiality. The new decision analyticsystem on the basis of Data Warehouse and Data Mining arises at the historicmoment.This thesis has introduced EDAS (Enterprise Data Analysis System)development and study on the current situation, and described the construction of theData Warehouse in detail from conceptual design, logic design, physics design;discussing data cluster analysis, related rule, decision tree algorithm analysis, thengoing on research, the practical data combined with the Data Warehouse platformhas carded on the following research: To a sales promotion, cluster analysis appearsto customer's consumption behavior, develop colony characteristic, businessoperating position of customer's colony and carry on detailed analysis to the market,enable the market department to adopt different marketing schemes in thedevelopment of business to different colony targets, thus realize enterprise's profit ismaximized; For realize real revenue and matching up to cost, adopt cluster algorithmand related rule going on amount charge matching, make cost of enterprise managedistinct, offer key data basis for policy-making level; To analyze the course ofsetting up decision tree, using decision tree technology to analyze customer's loss,analyze customer characteristic and relevant reason of loss, make correspondingtactics, offer decision basis for market managers and policy-makers to retrieve thecorresponding customers.