Volume 104, Number 2, October 2013
|Number of page(s)
|Interdisciplinary Physics and Related Areas of Science and Technology
|25 November 2013
Statistical properties of the personal social network in the Facebook
Research Center of Complex Systems Science, University of Shanghai for Science and Technology Shanghai 200093, PRC
Received: 23 May 2013
Accepted: 25 October 2013
The statistical properties of the user interaction behaviors in a city have great significance for developing the network marketing strategy, promoting personalized service and so on. In this paper, we investigate the interaction property of the users from New Orleans network in the Facebook, and find that one's out-degree and in-degree are approximately the same. In addition, when the number of a user friends is less than 65, the number of their posts would linearly grow with the slope 4.2, but when one user's friends are more than 65, their posts would grow with the slope 2.1. Further, the average link weight is relatively flat when the out-degree ranges from 28 to 65, and before or after the section it is on the rise or in decline, respectively, from which we can conclude that one could not maintain stable and meaningful relationships with more than 65 people in a single city. We present a null model to reshuffle the network to guarantee that the empirical results are not obtained by accident. The result obtained after reshuffling suggests that there exists a limit that restricts people's social activities.
PACS: 89.20.Hh – World Wide Web, Internet / 87.23.Ge – Dynamics of social systems / 05.10.-a – Computational methods in statistical physics and nonlinear dynamics
© EPLA, 2013
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.