计算机科学 ›› 2011, Vol. 38 ›› Issue (Z10): 370-375.

• 无线网络与通信技术 • 上一篇    下一篇

用SOM网络从移动通信数据中识别朋友关系

陈蔼样   

  1. (广东商学院数学与计算科学学院 广州 510320);(中山大学软件研究所 广州 610296)
  • 出版日期:2018-11-16 发布日期:2018-11-16
  • 基金资助:
    本文受广东省自然科学基金(104b103200100614D),广州市科技和信息化局(1OC1214D131),广东高校优秀青年创新人才培育项目(工YMl0081},广东省大学生创新实验项目(1069210097}资助。

Using SOM to Recognize Friendship from Mobile Phone Data

CHEN Ai-xiang   

  • Online:2018-11-16 Published:2018-11-16

摘要: 从移动电话通信数据中挖掘用户之间的有用信息被认为是富有研究价值的一项工作。利用从94个Nokia 6600用户中收集的实际通信记录,设计并实现了一个行之有效的高效分析算法do乙del,以构造手机号码对之间的通信模式,证明了do &.del算法的可靠性,并给出了算法的理论复杂性和实验观察结果。然后在do&del算法的基础上,给出了基于SOM(Sclf-Organizing Map)的关系识别系统RRS(Rclationship Recognition System),该系统能够在训练好的SOM网络支持下,有效识别双向朋友、单向/普通朋友、一般人等此类用户关系。

关键词: 现实数据挖掘,移动通信,SOM, Do&dol算法

Abstract: Data collected from mobile phone、have the potential to provide insight into the relational dynamics of individuals. In this paper, using the reality data collected from 94 Nokia 6600 users, an highly effective parsing algorithm named do&dol was designed and realized to construct the communication pattern of a pair phone number. The do&dol algorithm's correctness, theory time complexity and experimental observations were given. With the output of do&dol as training data, a Relationship Recognition System building on SOM neural network was designed and realized. The system RSS can recoginze the reciprocal friends, nonreciprocal friends and reciprocal nonfriends.

Key words: Reality mining, Mobile phone, SOM, Do&dol algorithm

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!