Design And Implementation Of Mobile Data Service Precision Marketing System For Operators

In face of user's diverse,personalized needs in the mobile Internet era,traditional marketing method of operators can not locate user demands accurately,blind marketing method is easy to cause the user's resentment and complaints,therefore,innovative marketing method is the important way that help operators to get rid of the fate of "Pipeline".In order to help operators to solve the problem which include massive data processing,user preference analysis and precision marketing scheme design,the design and implementation of the precision marketing system has become the urgent demand of operators currently.This paper is based on the "Ministry of Education-China Mobile" project.According to the needs of current operator's marketing,this paper put forward the operator-oriented mobile data service precision marketing system which include many key technologies,such as Hadoop platform for massive data processing,distributed crawler technology to collect user information,deep packet inspection technology to identify business,Hive and algorithm for user preferences analysis.The system is based on the concept of Hadoop distributed design,adopt the hierarchical and modular idea to meet requirements of low coupling,high performance and high reliability.The main work of this paper is as follows:Firstly,this paper put forward the operator-oriented mobile data service precision marketing system,which has high practical value and significance.Secondly,based on Hadoop distributed framework,this paper propose a distributed rape scheme,which overcomes the problem that the traditional reptile is low efficiency and is affected by the bandwidth easily.The reptile efficiency is 3 ~ 4 times higher than that of the traditional crawler.Thirdly,the traffic business identification module is divided into Primary DPI and Senior DPI,which can adapt to the change in demand better.Fourthly,the precise marketing module is divided into Hive-based marketing and Algorithm-based personalized marketing,which meet the evaluation of user preferences in different business environment and improve marketing success rate.At present,the system has been applied to actual marketing activities of operator in southwest successfully.The system is stable,good performance and provide strong support for marketing activities of operators.