The Design And Implementation Of The Data Warehouse System Of Vehicle Information Industry

With the rapid development of automobile industry, vehicle information industry also arises at the historic moment. At present domestic vehicle information industry is in the initial stage and has not formed a complete industrial chain. But part services of the vehicle information industry, such as vehicle navigation, have formed a certain market scale. At present, the main function of the data warehouse system in vehicle information industry is to query and simple online transaction processing. The business database accumulates large amounts of business data, and some of them are valuable information. While the traditional online transaction processing has not been able to meet the users' demands for data analysis,which is caused by the company has the massive data but do not find valuable information.Massive, meaningful data is stored in the database by common data. The result causes a serious waste of information.Based on the research of the data warehouse and data mining, deeply exploring is how to solve the internal system data inconsistency. How to ensure the daily data can be extracted to the reporting database per day. It is proposed in the overall design and implementation process of data warehouse and established the corresponding star model. Based on the data warehouse and data mining, it digs out the knowledge through the display system and provides a reliable basis for the formulation of marketing strategy management. This paper mainly includes the following aspects:First of all, the paper analyze the current status of vehicle information industry, theapplication of data warehouse and data mining technology and sort out the whole vehicle information industry business processes. This paper mainly introduces the process for vehicle registration and analyzes data features of the vehicle information industry with multi-source,type diversity, massive data characteristics. And ultimately determine the user needs through user communication many times, including the vehicle information query, the customer information query, the service information query, the order information query, the advisor information query, also providing the advisors clustering analysis function and the customer classification analysis function. The paper also contains the performance of the system, such as the system interaction, the network data flow and the data storage. The paper has the description of the vehicle information industry data warehouse system of the whole system structure, including the construction of data warehouse according to cleaned data, the data mining based on data warehouse and the data analyzing based on the data mining.Secondly, according to the business process and data characteristics of the vehicle information industry and the integration of user needs, the paper design the system architecture of the data warehouse system. Whatever the application of data warehouse and data mining technology, it cannot do without the large amounts of data. These data will be stored in the data warehouse after cleaned up, not only can obtain relevant information for the users, but also can be ready for data mining. This paper researches on data preprocessing in this process, mainly for different system data inconsistency of the same meaning by the signs.This paper uses the "top-down" approach to construct the data warehouse, and finally get the star model of data warehouse in the vehicle information industry.Then, based on the construction of the data warehouse, the advisors clustering analysis by using K means clustering algorithm and the customer classification analysis by using the C4.5decision tree algorithm, which use K means clustering algorithm to get the test results customer training set, and validate the model by using the cross validation method.Finally, the paper use the actual case build the data warehouse and realize the function of the system, the advisors clustering analysis and the customer classification analysis so as to obtain the corresponding management strategies such as the advisors are divided into 5 classes:the VIP group, the presentation(including registered vehicle) group, the emergency group, the sales group and the ordinary group. Each advisor is divided into different groups and they can play better features according to their different characteristics. The customers are divided into4 categories: the large customers, the important customers, the potential customers and the general customers. Each user are divided into different types of customers by different self-characteristics and behaviors and it can make marketing personnel better understanding of customer for which service they need and make the service better customer service.