Research Of The Data Mining In The Analysis Of Telecom Operates

With the reform and reorganization of China telecom system, the China telecommunications market environment has undergone fundamental changes. China's telecommunications service market gradually formed from individual operators to monopolize the market to a number of large operators of the new pattern of competition.The carriers operate as a business telecommunications service provider also has its own unique characteristics of customer and market: the telecommunications industry clients have diversity; telecommunication customers'demand have characteristics of diversity, showed great industry and the volume of business Features; telecommunication's services is more important than other industry's, telecommunications products, high renewal rate; In addition, the use of telecommunications services to customers is strong randomness, poor stability, users of telecom business operators have more rights to choose. In the face of a new, more intense, and a full range of international and domestic competition in the market environment, the domestic carriers is an urgent need for a powerful weapon which can provide the necessary support of the decision-making for telecommunications enterprise charges, marketing, planning and other aspects—Telecom Operations analysis and decision support system. This tool also must have the function of data mining.Data mining is a new and emerging technology,in recent years,accompanied by the development of artificial intelligence and database technology. Its core function is from the huge data sets or data warehouse to obtain useful information for business analysis and processing of complex data relationships. With the telecommunications market, increasing competition, operators generally began to "customer-driven" management model. In recent years, data mining technology with its powerful data analysis functions are widely applied to telecommunication carriers customer relationship management, marketing and other activities.This paper mainly studies the data mining in the analysis of telecom operates.After in-depth analysis of the way of constructing the Enterprise Data Warehouse,this paper offered a view of research of Data Mining in Telecommunication Industry,based on the predecessors.And provide an algorithm of the data mining in correlation analysis of customer call patterns especially for the specific characteristics of the telecommunications industry—Hash algorithm based on partition and Grid-based fuzzy clustering algorithm applied to the Customer Segmentation. Hash algorithm based on partition is a useful complement to the commonly used correlation analysis, achieved the design of the mass data partition, greatly reduced the number of database- scans,and successfully achieved the integration of hash technology and district technology, Not only shows the feasibility of the datamining-modular design, but also for improved the performance of the application of the various mining algorithms in telecommunications industry targeted to the characteristics of the telecommunications industry; Grid-based fuzzy clustering algorithm is the combination of Grid-based clustering and density-based clustering algorithm, realized in the applications of cluster to identify the characteristics of major clients in the telecommunications. Academic study of data mining in the massive are ascendant, and it hopes to provide valuable reference to the research of building and implementation of data warehouse in the telecommunications industry through the discussion of this paper.In addition, " The Law of Gravity " thinking in the nature will be introduced into grid-like data space model in this article.that is , first of all, the data space is divided into a finite number of unit grid structure, all the processing faceing to a single target grid, however ,those data objects near this cell and have "a closer relationship" to this grid will be attracted under the "gravity" principle, so that those data near the border of the cell was re-assigned to the cells which has " the largest attractiveness " to them. this is equivalent to blurring the boundaries of the cells. So as to have solved the problem of the traditional mesh clustering in which not considered the relationship between the adjacent unit,and have overcomed the shortcomings of the result- border of the cluster or level or vertical, and there is no border on the angle. At the same time eliminated the grid border too strong to some extent, she will have important guiding significance to the construction of the Analysis and Decision Support System of telecom operators,and will plays an important role in the process of building the actual system.