Home // ICIMP 2015, The Tenth International Conference on Internet Monitoring and Protection // View article

Multi-scaling Analysis of Social Network Users' Profiles

Authors:
Paulo Salvador
António Nogueira

Keywords: Facebook users activity; user profile; scalogram; cluster

Abstract:
Online social networks are extremely popular services on the Internet, having a relevant impact on the social, political and economical aspects of our daily lives. Being able to characterize social network users based on their temporal behavior and activity profiles can be useful to several engineering, management and design tasks. This paper characterizes the activity profiles of Facebook users, as well as their time evolution, based on data extracted using the Facebook Graph API and corresponding to social networking activity of 200 users during the years of 2011 and 2012. Using multiscale analysis based on a wavelet decomposition of the dataset, the proposed methodology is able to identify three main periodicity trends on the activity profiles: one of the identified profiles exhibits a visible periodicity around the 24 hours time scale (corresponding to users that have a daily interaction with Facebook), a second profile exhibits relevant components in the complete frequency range (intensive interaction) and a third profile includes important low periodicity components (users that tipically use Facebook in time periods higher than 24 hours). By looking at the temporal evolution of the different activity profiles, it is also possible to understand the dynamics of the profile changes.

Pages: 18 to 23

Copyright: Copyright (c) IARIA, 2015

Publication date: June 21, 2015

Published in: conference

ISSN: 2308-3980

ISBN: 978-1-61208-413-8

Location: Brussels, Belgium

Dates: from June 21, 2015 to June 26, 2015