基于URL分析的移动互联网用户分群

发布时间：2018-05-01 05:43

本文选题：移动互联网 + 网页分类URL　；参考：《河北大学》2013年硕士论文

【摘要】：近年来，移动互联网发展迅速。2012年7月，CNNIC公布的数据显示，60%的人通过移动设备接入互联网。移动互联网用户规模已不容小视，移动互联网市场潜力巨大。中国电信运营商非常重视移动互联网市场，密切关注着其可能带来的巨大商机。目前，各大运营商都在采取积极措施，分析用户的上网行为特征，从而找出目标客户群，根据客户群的不同向其提供差异化的服务，进而达到精准营销的目的。在此背景之下，本文提出并实现了一个基于URL的网页分类算法。该算法利用URL的特征频率进行网页分类，是一个具有较低计算复杂度的网页分类算法。应用该算法本文实现了一个基于URL分析的移动互联网用户分群系统，根据移动用户的上网行为轨迹实现了用户分群，通过设定营销规则，发现群组中符合营销规则的目标客户。本文的研究工作主要包括以下几方面：第一，本文研究了传统的网页分类算法和文本分类算法，在此基础上提出了一种基于URL的网页分类算法，该算法仅利用网页的URL特征对网页进行分类，算法主要特点就是计算复杂度低，，速度快。第二，根据某省移动公司的营销需求，设计了一个基于URL分析的移动互联网用户分群系统。该系统采用本文提出的基于URL的网页分类算法，对DPI日志中的URL进行分类，实现了移动互联网用户分群；本系统与移动公司的广告分发系统直接相连，可以实现对目标用户的精确营销。第三，本文通过一个营销实例验证了用户分群系统的功能，分析了系统的精确营销效率。
[Abstract]:In recent years, the mobile Internet has grown rapidly. Data released by CNNIC in July 2012 show that 60 percent of people connect to the Internet via mobile devices. The scale of mobile Internet users is not to be underestimated, mobile Internet market potential is huge. Chinese telecom operators attach great importance to the mobile Internet market and closely monitor its potential huge business opportunities. At present, all the major operators are taking active measures to analyze the characteristics of users' online behavior, so as to find out the target customer groups, and provide differentiated services to them according to the different customer groups, and then achieve the goal of accurate marketing. Under this background, this paper proposes and implements a web page classification algorithm based on URL. This algorithm uses the feature frequency of URL to classify web pages, and it is a low computational complexity web page classification algorithm. Using this algorithm, a mobile Internet user cluster system based on URL analysis is implemented. According to the mobile users' behavior trajectory, the users are divided into groups. By setting up marketing rules, the target customers in the group are found to conform to the marketing rules. The research work of this paper mainly includes the following aspects: Firstly, this paper studies the traditional web page classification algorithms and text classification algorithms, and then proposes a web page classification algorithm based on URL, which only uses the URL features of the web pages to classify the web pages. The main characteristic of the algorithm is its low computational complexity and high speed. Secondly, according to the marketing demand of a province mobile company, a mobile Internet user grouping system based on URL analysis is designed. This system adopts the URL based web page classification algorithm, classifies the URL in the DPI log, realizes the mobile Internet user group, and the system is directly connected with the advertisement distribution system of the mobile company. Can realize the accurate marketing to the target user. Thirdly, this paper verifies the function of user grouping system through a marketing example, and analyzes the accurate marketing efficiency of the system.
【学位授予单位】：河北大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP393.092

【参考文献】