Cluster Sowntown And Appliction Study Based On Least Cluster Cell

In order to win existence and development, most of the corporations have to make their production get partial resource in limited market and vehement competence. CRM system was bulit in many undertakings. As a integrate CRM system, there are four parts: information technology, operation technique, business intelligence, customer relation. Data warehouse, OLAP and data mining technologies have established a new framework of an enterprise decision support system, which enables DSS to enter into a phase of practical application. Cluster analysis has recently become a highly active topic in data mining research field. In the field of data mining, efforts have focused on finding methods for high performance and effective cluster analysis in large databases. Real-world applications may need to perform clustering under various kinds of constraints.In allusion to these questions, the author of this paper have studied over constraints and technique of cluster analysis. In order to find methods for efficient and effective cluster analysis in large database, this paper probes sowntown cluster that is base on least cluster cell by the analysis of cluster methods. The main thought is to introduce LCC managing subsystem in the general data-mining system architecture. The mining results, that is LCC, which were provided to the users directly in the past, are stored in the LCC database through which the users can carry on various operations and management. A consistent view of LCC is formed so as to improve the efficiency of sharing and mining cluster sowntown. After studying various constraints, a constraint-based cluster using function partition based on angle dissimilarity, a partition sowntown means with least outspread tree method, and the quantity of customer's sale was forecasted by Time-Series Analysis were presented.. So, a series of CSPR(Cluster Sowntown Partition Rule) to manage LCC database was describe. The CSPR based on various kinds of constraints was advanced, such as partition rules of areas , divisionrule of piece, layout rules of deliver goods line. By CSPR, Java, ODBC and IBM DB2, the author carries out the pivotal technology and major element. At last, this paper makes an experimental system for sowntown cluster and multidimensional analysis of YUZHONG TOBACCO company. This corp. put MIS in practice in 1996, so there are lots of data in data warehouse. And because of the complex geographic environment and human culture, such as city's rivers and bridge, lane, the direction of hill, and customer sales per region, this experiment have some representation. The success of this experiment proved it is viable of the sowntown cluster based on LCC. There are some theoretical and practical values in similarity problem of commerce, finance and geological exploration.There are eight chapter in this thesis. It included:1 Exordium, 2 Summarize of bulit data warehouse,3 Summarize of data mining, 4 Summarize of cluster analysis, 5 Least cluster cell (LCC), 6 Sowntown cluster arithmetic with constraint based on LCC, 7 Sowntown cluster arithmetic with constraint based on LCC was carried out in BIMINER, 8 The design and analysis of sowntown cluster experiment in BIMINER.