电信科学 ›› 2017, Vol. 33 ›› Issue (7): 112-119.doi: 10.11959/j.issn.1000?0801.2017127

• 研究与开发 • 上一篇    下一篇

一种识别骚扰电话的组合算法研究

王彦青1,王瀚辰1,2   

  1. 1 东信北邮信息技术有限公司,北京 100191
    2 美国特尔菲学校,美国 俄勒冈州 谢里登 97378
  • 修回日期:2017-03-15 出版日期:2017-07-01 发布日期:2017-07-25
  • 作者简介:王彦青,男,东信北邮信息技术有限公司数据分析师,主要从事算法研究方面的工作。|王瀚辰,男,现就职于东信北邮信息技术有限公司。

Research on a combining algorithm for harassing calls to identify

Yanqing WANG1,Hanchen WANG1,2   

  1. 1 Eastcom-BUPT Information Technology Co.,Ltd.,Beijing 100191,China
    2 Delphian School,Sheridan,OR 97378,USA
  • Revised:2017-03-15 Online:2017-07-01 Published:2017-07-25

摘要:

当前骚扰电话层出不穷,严重影响了人们的日常生活。为了有效防范此类电话带来的不良社会影响,采用了数据挖掘的分析手段,深入研究了骚扰电话呼叫特点,提出了一种基于用户反馈的分时段分析骚扰电话识别方法;并对用户标识的疑似骚扰号码引入随机森林算法,极大提高骚扰源识别率,结合布控拦截机制,整体实现对骚扰电话全方位的管控。通过实际数据验证,效果明显。

关键词: 骚扰电话, 随机森林算法, 大数据

Abstract:

At present,people’s daily life has been seriously affected by an endless stream of harassing calls.To prevent the adverse social influence,a harassing calls recognition method based on the analysis of users’ feedbacks was proposed,which could make people look insight into the features of harassing calls by data mining.Also,the random forest algorithm was applied to identify the suspected harassment numbers.In this way,the recognition rate of harassment source has been enhanced greatly,and a comprehensive control in harassing calls can be achieved by integrating the portable interceptor.Simulation results also show its good performance.

Key words: harassing call, random forest algorithm, big data

中图分类号: 

No Suggested Reading articles found!