The Application And Research Of Data Mining System In ISBN Management

Since the ISBN real-name application information system has being used in nationwide of china, there have bean one million information of books, while the system offer convenient service for ISBN also exposed irregularities of publishers in the process of book publishing, for example, violation publication, the information of ISBN repeatedly revised, books terminated too many times, the integral information of ISBN can not be uploaded. In addition, the regulatory agencies of publishers wants to summed up the rule of the current publishing book from one million existing information of published books, in order to better grasp the overall trend of the book market. Based on the above requirements, this paper lead the thinking of data mining to the field of book publishing, and it designed and implemented the mining system of ISBN.This paper firstly studied on the background of areas related to the topic and the current situation. Secondly, from the book publishing industry problems and the business needs of the administration, and combined with the technology of data mining, it proposed the models of data mining based on the subject field. In this paper, the early initial time of creating data warehouse, in order to migrate this problem of solve data between heterogeneous databases, it proposed a JDBC-based heterogeneous database migration, and gave a good solution of data migration between different database systems and different versions of the database system.In the data pre-processing stage, from studying of business processes of the ISBN real-name application information system, and combined with the techniques of data mining, it proposed two kinds models of data mining:firstly, the subject domain association rules model based on the distribution of types of books, and the use of classic Apriori association rules analysis algorithm for the type of books published by the national585book publishers association rules analysis to identify a period of time hot spot, and book publishing book publishing industry, the overall trend; publishing publishing House Books behavior specification of the subject domain clustering model, modify the number of times a Press book publishing process books, the number of termination, a book is not the number of upload, upload extended number of times, the use of the classic K-means clustering algorithm for clustering analysis, identify problems and prominent publishing house collection management basis for the publishing house managers.In this paper, the ISBN book data mining system has been developed in the book publishing industry, can effectively regulate the book publishing behavior of publishers, and there is of great significance for china’s publishing industry to healthy and orderly developed.