移动互联网的用户行为分析系统的设计与实现
发布时间：2018-10-30 15:08
【摘要】： 2008年,我国开始3G的商用,据统计,移动互联网用户每天的增速正在超过传统互联网用户；传统电信运营商与用户之间单向的业务收费的运营模式转变成了以用户为中心的多元盈利模式,运营商的竞争对手来自互联网,来自终端厂商,运营商要在新的通信行业的发展下保持领导地位,不能仅仅作为通道,必须应用自己已有的资源向用户提供新增长点的内容与业务,那么如何和用户提供高附加值的业务?其中一个热点的方向就在于对移动用户的行为进行分析,得到用户的行为特征与分类,在保护用户隐私的前提下为之提供精准的目标服务,从而在提供带宽的同时,为用户提供高附加值的目标服务,实现运营商与用户的双赢。 在此背景之下本论文设计一个基于移动互联网的用户行为分析系统,系统位于中国移动或者中国联通的核心网分组域中的SGSN与GGSN之间Gn口或者GGSN与分组IP网之间的Gi口,处于封闭式或者半开放式网络当中,在安全性上较为有保证；系统选取用户具有代表性的WEB浏览行为作为研究对象,对用户数据进行采集,加以过滤,运用DPI(深度包检测)的思想对数据内容进行分析,采用数据挖掘中的支持向量机(SVM)算法对内容进行分类,得到用户行为特征与分类；此外,用户行为分析系统可以与运营商的业务生成模块直接相连,具体来说,与目标广告分发子系统相连,从而实现可能的精准目标广告。 论文首先分析了3G时代电信运营商面对的全新的业务挑战的背景,提出了基于移动互联网的用户行为分析系统的系统框图。系统由数据采集模块,数据DPI模块,内容分类模块,用户特征数据库等模块组成；对系统中的数据DPI模块,内容分类模块,用户特征数据库进行了详细设计。最后对本系统进行实验环境搭建,给出实验结果与分析。最后总结了论文的特点以及不足之处,并提出了对前景的展望。
[Abstract]:In 2008, China began 3G commercial, according to statistics, mobile Internet users are increasing faster than traditional Internet users every day; The traditional operating mode of one-way service charges between telecom operators and users has changed into a multi-profit model centered on users. The competitors of operators come from the Internet and from terminal manufacturers. In order to maintain the leading position under the development of the new communication industry, operators must use their existing resources to provide users with new growth points of content and services, so how to provide high value-added services with users? One of the hot spots is to analyze the behavior of mobile users, get the behavior characteristics and classification of users, and provide accurate target services for mobile users under the premise of protecting their privacy, so as to provide bandwidth at the same time. To provide users with high value-added target services, operators and users to achieve a win-win. Under this background, this paper designs a user behavior analysis system based on mobile internet. The system is located in the Gn port between SGSN and GGSN in the core network packet domain of China Mobile or China Unicom or the Gi port between GGSN and IP network. In the closed or semi-open network, the security is guaranteed; The system selects the representative WEB browsing behavior of the user as the research object, collects and filters the user data, and analyzes the data content by using the idea of DPI (depth packet Detection). The support vector machine (SVM) algorithm in data mining is used to classify the content, and the user behavior features and classification are obtained. In addition, the user behavior analysis system can be directly connected to the service generation module of the operator, specifically to the target advertising distribution subsystem, so as to achieve the possible accurate target advertising. Firstly, the paper analyzes the background of the new service challenges faced by telecom operators in 3G era, and puts forward the system block diagram of user behavior analysis system based on mobile Internet. The system is composed of data acquisition module, data DPI module, content classification module, user feature database module, etc. The data DPI module, content classification module and user feature database in the system are designed in detail. Finally, the experimental environment of the system is built, and the experimental results and analysis are given. Finally, the paper summarizes the characteristics and shortcomings of the paper, and puts forward the prospect of the future.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2010
【分类号】：TP393.01;TN929.5
本文编号：2300345
[Abstract]:In 2008, China began 3G commercial, according to statistics, mobile Internet users are increasing faster than traditional Internet users every day; The traditional operating mode of one-way service charges between telecom operators and users has changed into a multi-profit model centered on users. The competitors of operators come from the Internet and from terminal manufacturers. In order to maintain the leading position under the development of the new communication industry, operators must use their existing resources to provide users with new growth points of content and services, so how to provide high value-added services with users? One of the hot spots is to analyze the behavior of mobile users, get the behavior characteristics and classification of users, and provide accurate target services for mobile users under the premise of protecting their privacy, so as to provide bandwidth at the same time. To provide users with high value-added target services, operators and users to achieve a win-win. Under this background, this paper designs a user behavior analysis system based on mobile internet. The system is located in the Gn port between SGSN and GGSN in the core network packet domain of China Mobile or China Unicom or the Gi port between GGSN and IP network. In the closed or semi-open network, the security is guaranteed; The system selects the representative WEB browsing behavior of the user as the research object, collects and filters the user data, and analyzes the data content by using the idea of DPI (depth packet Detection). The support vector machine (SVM) algorithm in data mining is used to classify the content, and the user behavior features and classification are obtained. In addition, the user behavior analysis system can be directly connected to the service generation module of the operator, specifically to the target advertising distribution subsystem, so as to achieve the possible accurate target advertising. Firstly, the paper analyzes the background of the new service challenges faced by telecom operators in 3G era, and puts forward the system block diagram of user behavior analysis system based on mobile Internet. The system is composed of data acquisition module, data DPI module, content classification module, user feature database module, etc. The data DPI module, content classification module and user feature database in the system are designed in detail. Finally, the experimental environment of the system is built, and the experimental results and analysis are given. Finally, the paper summarizes the characteristics and shortcomings of the paper, and puts forward the prospect of the future.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2010
【分类号】：TP393.01;TN929.5
【引证文献】
相关硕士学位论文 前2条
1 郭静;移动信息采集分析软件的设计与实现[D];解放军信息工程大学;2012年
2 孙新;融合空间谓词的高空间分辨率遥感图像关联分类方法研究[D];南京师范大学;2013年
，本文编号：2300345
本文链接：https://www.wllwen.com/wenyilunwen/guanggaoshejilunwen/2300345.html
