当前位置:主页 > 经济论文 > 房地产论文 >

基于朴素贝叶斯的移动客户职业识别算法与调试

发布时间:2018-03-27 12:34

  本文选题:朴素贝叶斯模型 切入点:移动用户职业 出处:《内蒙古大学》2013年硕士论文


【摘要】:随着移动通信、互联网业务的迅速发展。手机已经从奢侈品变成了生活日用品,是我们日常生活中不可缺少的一部分。人们随时随地使用手机打电话、发短信、上网,而用户的这些行为以及其个人基本信息均在运营商处有所记录。如何能够通过流量费、通话费、短信条数、上网时长四个加权因素来综合刻画用户职业类型。对于整体运营商行业来说,是极其重要和有价值的研究课题。 对特定人群提供特定服务和合适的资费尤为重要。因为不同的职业有不同的消费倾向和消费能力,对选择资费和话费业务上有很大的不同。所以通过用户以往数据,归纳出不同职业用户的特征,再对未知职业的用户做出预测,对公司市场和业务流程有很大的帮助,可以节约大量不必要的推广成本,提高推广转化率。 我们运用朴素贝叶斯模型先对信息完整的用户职业信息进行分类。以职场新人为例,根据刻画职场新人的四个影响因素与职场新人职业之间的关系。预测出两个职业信息不详的职场新人分别是从事房地产经营产业和其他商务服务业的。我们将数据表中手机信息不详的职场新人单独列举出来,根据上问中信息完整的职场新人的相关指标因子来确定手机信息不详的职场新人终端大致的价格范围。当职场新人与聚类中心的距离小的时候,这个职场新人就属于这个类别。当这个距离大的时候,该职场新人就不属于这个类别。
[Abstract]:With the rapid development of mobile communication and Internet business, mobile phone has changed from luxury to daily necessities, which is an indispensable part of our daily life. And the user's behavior and their personal basic information are recorded at the operator's office. How can they be recorded through traffic charges, phone calls, text messages, It is an extremely important and valuable research topic for the whole operator industry to describe the occupation types of users synthetically by the four weighted factors of the length of the Internet. It is particularly important to provide specific services and appropriate rates to specific groups of people. Because different occupations have different propensity and ability to consume, they are very different in selecting tariff and telephone services. The characteristics of different professional users are summed up, and then the users of unknown occupations are predicted, which is of great help to the market and business process of the company, which can save a lot of unnecessary promotion costs and increase the extension conversion rate. We use naive Bayes model to classify the occupational information of users with complete information. According to the relationship between the four factors that affect the new career and the occupation of the new workplace, we predict that the two new professionals with unknown information are engaged in the real estate business and other business services respectively. We will count. According to the list of new workplace workers whose mobile phone information is not available, According to the relevant index factors of the newlyweds with complete information in the above question, to determine the approximate price range of the new workplace newcomers with unknown mobile phone information. When the distance between the workplace newlyweds and the cluster center is small, This new worker falls into this category. When the distance is large, the new worker does not fall into this category.
【学位授予单位】:内蒙古大学
【学位级别】:硕士
【学位授予年份】:2013
【分类号】:TP311.13

【参考文献】

相关期刊论文 前8条

1 吴胜浩,钟亦平,张世永;消息总线的设计及其在电信领域的应用[J];计算机工程;2004年18期

2 马华;李建华;;基于SOA的企业应用集成系统[J];计算技术与自动化;2005年04期

3 杨涛,刘锦德;Web Services技术综述——一种面向服务的分布式计算模式[J];计算机应用;2004年08期

4 卢致杰,覃正;SOA构架与电子商务应用集成[J];计算机应用研究;2004年10期

5 张雪洁,王志坚,许峰,李静燕;基于XML的领域异构数据库间的数据转换[J];计算机与现代化;2004年05期

6 张月琴;;网格计算的安全性分析[J];软件导刊;2006年19期

7 周珂;基于SOA软件架构的企业应用[J];微机发展;2005年11期

8 孙国刚,范学峰,刘卫国;基于WS-Security和角色访问控制的Web服务安全性研究[J];微计算机应用;2005年01期



本文编号:1671550

资料下载
论文发表

本文链接:https://www.wllwen.com/jingjilunwen/fangdichanjingjilunwen/1671550.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户c9c45***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com