Geography Reference
In-Depth Information
means, OSNs clearly differ from other web-based networks like Internet infrastruc-
ture. The latter are led by power-law tie-distribution: a small share of webpages
accounts for an outstandingly high number of links (Barab
´
si and Albert
1999
).
In our understanding OSNs are “biased versions of real-life networks”
(Backstrom et al.
2012
; Ugander et al.
2011
). Therefore, we claim that virtual
space and physical world are strongly interrelated, since it is assumed that flesh and
blood users document their offline friendships in the online environment. The
linkage between virtual and physical spheres, namely geolocalisation of online
content is typically based on the position of users that can either stem from
voluntary geographic information that users attach to the content they upload
(e.g. image posts) or can be based on IP addresses etc. The possible projection of
virtual world on real geographies and the effect of location on shaping cyberplace
poses several questions that are beyond the subject of this paper. For example,
extending research is still missing on the correlation among online and offline social
networks (Traud et al.
2008
; Hogan
2009
). We only use the statement that geo-
graphical location and physical distance are very important in OSNs.
In recent years there has been a growing public and scientific interest in
analyzing OSNs; the mainstream of research covers a very wide area including
learning- and communication processes (Greenhow
2011
), online identity (Zhao
et al.
2008
), youth and digital media (Boyd
2008
), online privacy (Acquisti and
Gross
2009
), network dynamics (Kumar et al.
2006
), among others. Geography and
geovisualisation has been also involved to the discussion, for example in the field of
user-generated information mapping (Yardi and Boyd
2010
; Graham and Zook
2011
), infographic geovisualisation of network connections (Buttler
2010
; Warden
2010
) or in cartographic explorations of online and offline geographic relations
(G
¨
tz
2010
). Additionally, interesting experiments could be mentioned in connec-
tion with user behaviour mapping (Fischer
2011
) or geotagged information map-
ping (Graham and Gaffney
2012
) as well.
Research works on basic OSN characteristics highlighted that geography matters
in network formation. Liben-Nowell et al. (
2005
) stressed that only one-third of
friendships realized on LiveJournal blogging SNS was independent of bounded
geographical areas. Escher (
2007
) also found that majority of ego-networks are
local. A mega-analysis of Facebook found that majority of connections are within
country borders and the number of ties across countries accords with geographical
distance (Ugander et al.
2011
). Thus, the geographical location of a user strongly
determines the geographical position of the friendship ties he/she documents on the
website.
It came to the light also that distance seems to be of crucial importance in OSNs.
For example, research on Facebook, LiveJournal and Twitter also found evidence
on “small world phenomenon” (Backstrom et al.
2012
; Liben-Nowell et al.
2005
;
Yardi and Boyd
2010
). Users formulate strongly connected cliques with physically
proximate other users whereas relatively few long distance ties make the whole
network connected and establish short average paths between two random users.
There might be differences among distance types as well; for example Takhteyev
et al. (
2012
) found that the frequency of airlines between two cities has the strongest
