The Design And Implementation Of Data Management Platform For Telecom Industry Under The Environment Of Big Data

The development of the internet brings the impact to the development of traditional enterprise,not only including the model of the traditional entity store line sales impacted by the electricity supplier,but also the virtual telecom operators impacted by the traditional telecom operators.The traditional operators need to make the corresponding transition to adapt to the trend of the times.The telecom operators must learn to solve the problem of how to deal with the massive fragmentation of data,and how to improve operational efficiency to meet the growing needs of individual users.This thesis attempts to design a set of data platform for the telecommunications industry to simplify data processing of massive and fragmentation data to help staff to develop precision marketing solutions in order to improve operational efficiency.This thesis does the following works based on being familiar with data mining techniques and theoretical knowledge,the analysis of the development direction and the data characteristics of the telecommunications industry,and having mastered the technologies of web application platform development and data storage.Firstly,with the use of Node.js web development,this thesis finishes the design and implementation of a web application of data management platform to show the telecom data mining valid intuitive results,and tries to use the friendly and interactive experience to reduce technical barriers related to the work of the telecom data mining.Secondly,it is from the understanding of the type and the characteristics of telecom operator's subscriber data,product data and historical marketing data that it develops a suitable data analysis program of data traffic business,and constructs a model of customer segmentation based on data traffic service.Thirdly,for the storage of telecom data,the platform builds the cluster environment of Hadoop distributed storage system,completes the data storage by using a distributed file system HDFS,and finishes the query and analysis of data by using a distributed data warehouse Hive.Fourthly,for the analysis of data processing,it uses the R language server program Rserve to implement the relevant algorithms on the Hive data processing,and transfers the results to Web with the RIO library.With the use of the Web application technologies Node.js that supports the real-time and two-way connection establishing station,and it uses Rserve as back-end server to do massive data-processing,a distributed storage system Hadoop to finish data storage,this paper presents a set solution of telecommunications data management platform.The program offers excellence performance and has very high viscosity that can be easily associated with other mainstream technology to do combination and expansion.Finally,this thesis has achieved the result of data management platform by actual coding of each module,and after testing,the platform can perform the large data storage and analysis,and can be well treated for concurrency,and stable operation.