Using multi-features to recommend friends on location-based social networks

被引：0
作者：
Gao Xu-Rui
Wang Li
Wu Wei-Li
机构：
[1] Taiyuan University of Technology,College of Computer Science and Technology
[2] University of Texas,Department of Computer Science
来源：
Peer-to-Peer Networking and Applications | 2017年 / 10卷
关键词：
Location-based social network; Recommend friends; Multi-features; SVM;
D O I：
暂无
中图分类号：
学科分类号：
摘要：
Location-based social networks (LBSNs) not only offer novel services but also produce more abundant data to help new serves for human. It will help discover latent trajectory, possible friendship and then guide trip, predict next place, recommend friends, and promote sales and so on. In this paper we study two problems for friend recommendation on LBSN: what is the main feature and how to predict friendship? We firstly analyze many factors related with human mobility and social relations; adopt the information gain to measure the contribution of different features to human friendship. Then we extract user social relationship, check-in distance during fixed periods and check-in type as key features. Because the prediction problem could be considered as a classification problem, we choose SVM to predict friendship. At last some experiment results show our algorithm valid to some extent.
引用
收藏
页码：1323 / 1330
页数：7
相关论文
共 20 条
  • [1] Aiello LM(2012)Friendship prediction and homophily in social media ACM Trans Web (TWEB) 6 9-168
  • [2] Barrat A(2013)Friendship prediction in recommender system Journal of National University of Defense Technology 35 163-22441
  • [3] Schifanella R(2010)Inferring social ties from geographic coincidences Proceedings of the National Academy of Sciences 107 22436-11628
  • [4] Ma J(2005)Geographic routing in social networks Proceedings of the National Academy of Sciences of the United States of America 102 11623-1221
  • [5] Xu H(2006)Pattern recognition and machine learning Company New York NY 16 049901-C874
  • [6] Chen H(1998)Reinforcement learning: an introduction IEEE Transactions on Neural Networks 9 1054-36
  • [7] Crandall DJ(2012)Using Location-Based Social Networks to Validate Human Mobility and Relationships Models [J] IEEE 40 1215-undefined
  • [8] Backstrom L(2006)An introduction to ROC analysis Pattern Recognition Letters 27 861-undefined
  • [9] Cosley D(1982)The meaning and use of the area under a receiver operating characteristic (ROC) curve Radiology 143 29-undefined
  • [10] Liben-Nowell D(undefined)undefined undefined undefined undefined-undefined
12