基于漏话保护系统的用户数据分析

发布时间：2018-01-13 07:29

本文关键词：基于漏话保护系统的用户数据分析　出处：《南京邮电大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着电信行业的发展,各大运营商之间的竞争越来越激烈,同时移动通信的发展也影响到固定电话运营商,所以固话运营商纷纷采用各种措施来吸引用户,希望达到增加收入的目的,但是真正有效的方法是提供满意的服务,让用户体验达到较高的水平。漏话保护系统,是一个为了减少固话用户漏话而提出的增值业务系统,通过将漏话语音接入漏话保护平台,为用户提供友好的语音交互,可以实现为用户继续呼叫、短信通知等功能。这就表明该业务与以往的漏话短信提醒有所区别,所以该业务应该对哪些用户开放,针对不同用户采取哪些不同的策略,这是本文研究的重点。本文主要对漏话保护系统进行研究,包括业务背景、系统结构、业务流程等。漏话保护系统主要是针对政企用户,所以面临的最主要的问题就是用户选择,本文采用数据挖掘技术中聚类挖掘方法对用户数据进行分析,对用户进行分类,然后对每类用户进行定义,为了运营商能够对各类用户提供更好地个性化服务。本文的主要工作内容如下:(1)对漏话保护系统进行结构设计,分析业务背景和业务流程,找出漏话保护系统的主要问题:用户数据分析。(2)学习数据挖掘知识,并深入研究聚类挖掘方法,研究聚类挖掘方法中的K-means算法,并分析该算法的缺点。(3)对原K-means算法进行改进,采用SPSS Clementine数据挖掘软件,利用改进的K-means算法对漏话保护系统的用户数据进行建模,并将模型应用于数据分类。(4)对漏话保护系统的用户数据分类过程和结果进行展示,并对分类结果进行定义和分析,最后针对各类用户提出建议性的营销策略。
[Abstract]:With the development of telecommunications industry, the competition between major operators is becoming more and more fierce, and the development of mobile communication also affects fixed telephone operators, so fixed-line operators have adopted a variety of measures to attract users. To achieve the goal of increasing revenue, but the real effective way is to provide satisfactory services, so that the user experience to a higher level. It is a value-added service system which is put forward in order to reduce the leakage of fixed telephone users. It can provide users with friendly voice interaction and can continue to call for the users by connecting the missing voice to the leaky voice protection platform. This indicates that the service is different from the past short message alerts, so the service should be open to which users and what different strategies should be taken for different users. This is the focus of this paper. This paper mainly studies the leak protection system, including the business background, system structure, business process and so on. So the most important problem is the choice of users. In this paper, we use clustering mining technology to analyze the user data, classify users, and define each class of users. In order for the operators to provide better personalized service to all kinds of users. The main work of this paper is as follows: 1) the structure of the leak protection system is designed, and the business background and business process are analyzed. Find out the main problem of the leaky voice protection system: user data analysis. 2) learn the knowledge of data mining, and further study the clustering mining method, research the K-means algorithm in the clustering mining method. And analyze the shortcomings of the algorithm. 3) to improve the original K-means algorithm, using SPSS Clementine data mining software. The improved K-means algorithm is used to model the user data of the leaky voice protection system, and the model is applied to the classification of data. The classification results are defined and analyzed. Finally, suggestions are put forward for all kinds of users.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN929.5

【相似文献】