Research And Realization Of The Book-sale Intelligence Analyzes System Based On Multi-strategy Data Mining Models

Data mining involves a lot of academics; it includes database technique, statistics, the data visualization, machine study, artificial intelligence and high performance algorithm. With the fast development of data mining, he research emphases to data mining turned it step to engineering gradually from fundamental research. The actual application of data mining have already obtained huge achievement in the business intelligence, and make the business enterprise converting into the knowledge various complicated and huge data, then help a business enterprise to carry on management decision, even have the function of business intelligence decision.This paper with the beginning from the data of the book-sale, used the technology of data mining and the method of data visualization, also used mathematics and the technology related the discipline of programming language for computer. The primary coverage of this paper could summarize the following points:1. This paper introduced research background and basic concept of data mining technical, function, task, techniques used commonly of data mining and the development of future.2. This paper aims at different analysis function respectively and adopted using bayes, relational analysis, decision tree analysis, cluster analysis and so on to the data mining and carries on analysis to information of register and books.3. This paper chooses Microsoft of Microsoft SQL Server 2005 Analysis Servers(SSAS, SQL Server analyzes service) as data mining tool.4. Carry to pass ADODM.NET to visit SSAS database object at the customer and adjust to use DMX a search order to carry out a model estimate.Be analyzed enunciation by research, the application of data mining technique to book selling',can mine worthy knowledge from these complicated information, it also can provide effective decision support to the head of each department of business enterprise for book.Concretely is following points:1. Mining the registration member's information of bookstore, the customer supervisor could predict the member's Class and consumption type, then carry on an automatic classification to the member according to the member's personal information and consumption record, and recommend other possible interested in books to he/she according to the member's consumption record.2. Mining the information of each kind of books, the selling supervisor can predict future sale amount of money according to the information of the past sale of books and predict its sale quantity according to the book basic information, and estimate that book can bind other books of settling the sale, in the meantime according to book the basic information estimate list price to that book...