Car Sales System Data Warehouse Solution

At present, use and concentrate in the fields of trades , such as telecommunication , bank , securities ,etc. in the domestic data warehouse . In automobile sale trade, data warehouse application of technology popularization very much still. This subject applies the warehouse technology of the data to the decision analyticl system of the automobile sale to the need of enterprise's development, make every effort to offer a feasible solution. This data warehouse model which we attempt to finish, can be divided into ETL (data collecting (Extract ) , change (Transform ) , wash (Cleansing ) , the course loaded (Load ). ),DW (data warehouse ), three parts of DM (the data are excavated ). This text has probed into the business demands of enterprises again at first, then analyse about initial data, it is mainly that relevant database does not meet the needs of enterprise's business development , set up data warehouse , combine OLAP analyse technology and data excavate technology , is it favorable to enterprise knowledge message of development to find of a large amount of data, it guarantees to be convenient to visit and by good way answer complicated question fast to business datum at the same time.The demand is analysed have already confirmed that analyses the data needed in customer service. Will confirm the blueprint in the future of warehouse system of the data on design phase in model. The main task at this stage is that the logic of carrying on the data warehouse is designed, including choose the suitable theme , confirm fact form , linking relevant, attribute and grain size divide, design correct form structure and major key , other key relation ,etc. Step that the model is mainly designed including four are basic: Confirm the suitable theme , level of grain size of division , designing and linking forms and designing the estimation of the fact form , data quantity, etc..The pretreatment of the data, build one of the indispensable steps in the data warehouse too. While the data collect, wash , reprint, the quality problems of the data are shown especially out. Examine and put the data in order, guarantee the consistency of the data in the data warehouse , gather operating to some data , will raise the inquiry of the data warehouse , speed of analysis, and the comprehensive data can not enter the data warehouse in intact, unanimous, the detail of these. Key element on the question of quality of the data of this text, have classified and explained , illustrate the existing problem in data quality in this subject initial data with examples . The data that have analysed , commonly more used wash algorithms, and on the question of the repeated record in the database, to arranging in an order the shortcoming of neighbour's algorithm , in cluster's analysis algorithm , adopt Canopy technology, have reduce the calculating amount in the cleaning process, has reduced the complexity of the algorithm.The thesis still combines several kinds of daily data in the systematic comparative analysis data warehouse of automobile sale decision and stores the structure. Utilize 2000 of Analysis Services' multidimensions inquire language MDX is it inquire about to make , go on OLAP is analysed briefly. It is to the summary ofthis subject finally, and look forward to the working direction in the future.