Towards Multi-account Users Detection and Connection Based on Telecom Data

DOI

10.2991/icmit-16.2016.35How to use a DOI?

Keywords

DPI, connecting users, similarity, user profile, behavioral similarity, co-occurrence

Abstract

Internet users tend to have a number of different types of Internet accounts, and yet there does not exist a perfect Internet account association approach. The telecom network operators use DPI technology can detect the part of user's Internet records, which makes it possible to realize the connection of the Internet account on the operator side. This paper illustrates the significance of extraction, connection of Internet accounts from the DPI data. We combine the edit distance and Jaccard coefficient two string similarity algorithms to propose a string similarity method which is suitable for calculating the similarity of the account information and use classification algorithm to analyze the similarity of the account from the user's behavior, and find the correlation between the accounts. On the basis of the account associated, we provide more detailed portrait of users. Finally, we show that the proposed method can effectively extract the accounts from DPI data and associate accounts experiment on China Telecom data.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Zhongshun Hu
AU  - Fei Chen
AU  - Chenhao Xie
AU  - Yanghua Xiao
PY  - 2016/04
DA  - 2016/04
TI  - Towards Multi-account Users Detection and Connection Based on Telecom Data
BT  - Proceedings of the 2016 3rd International Conference on Mechatronics and Information Technology
PB  - Atlantis Press
SP  - 195
EP  - 199
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmit-16.2016.35
DO  - 10.2991/icmit-16.2016.35
ID  - Hu2016/04
ER  -