Design And Implementation Of A Platform For Telecom Data Processing Based On Cloud Environment

Data is the lifeblood of the modern information society, but also is the key of bottlenecks of Telecom service providers layout data services industry. Due to the competition of telecom data services sectors, Data gathering, Data mining has become a extraction and the core of decision support, provide knowledge and played an important role. Due to the bussiness traditional’ information way, the traditional query, statistic technology only can be for the basic data processing, difficult to analyse the data on the higher level. In the data gathering, Processing power on the different degree restricts the service level.Now I’m serving in the China Telecom Besttone Ltd(hereinafter referred to as: Besttone), with the gradually rise of the requirement of data quality and data management, It need a new platform to solve quality problem in the traditional telecom system data acquisition and data interference problems, and on the basis of the original data integration improve data processing depth, improve the accuracy of data mining, improve data processing efficiency, reduce the data processing unit cost.This paper shows that the cloud computing environment design and implementation of data processing platform of Telecom, the main contents include the following aspects:(1) The thesis research background: As a telecom professional data management company, there is a shortage of huge amounts of data with the data processing ability of the contradiction.As a result, we need established a cost-effective and efficient data processing platform based on the cloud computing environment to support the company’s all aspects of business conduct.(2) The key technology introduction: The paper introduces the basic theory and key technology of data processing platform application, including cloud computing and the relevant concepts of data mining and common mining algorithm. Introduced some real cases of data mining and cloud computing,and part of application situations of Cloud Computing at home and abroad,explain the necessity and superiority of cloud computing framework.(3) Research of platform requirements and algorithms: The platform project requirements and architecture selection process. Paper illustrates the characteristic of data processing and expounds data processing platform project requirements of our company. The paper expounds the platform selection based on the necessity of Apache Hadoop cloud computing architecture. At the same time, the paper discussed these implementations in cloud computing environment: Customer classification, Customer repetitive judgment, Non-canonical text processing, Information extraction, and other specific function algorithm. the paper also shows the data process of the telecom log type data and the base type data.(4) The platform design and implementation: The paper illustrates the design of the platform based on use case view, logical view, system technology platform view. Introduced Whole platform model. Introduced platform of three main modules(front desk service module, the parallel transaction management module, Hadoop service module).(5) Platform test and validation: the paper elaborate on test environment and test tasks, and analysis of test results to validate the actual usage,and check the performance indicators of the verification platform.(6) Summary and prospect: Data processing platform has been successfully applied after online, obtained the actual benefit. Platform provide strong support for the company in providing Segmentation analysis, Classification analysis, Customers repeat judge.