[Air-L] facebook and ethnicity

[Seda Guerses]

hello again,
thank you for all the comments.
i was very curious about the facebook study, also because of the names  
of the researchers who executed the story. lars backstorm has been one  
of the first authors to tackle the problem of impoosibility of  
"anonymity" in social network like data structures ***. his paper  
together with cynthia dwork (the person who developed differential  
privacy) and jon kleinberg (a prominent social network/graph analyst)  
was one of the first to show that simple anonymization of a social  
network by simple de-identification of nodes in the network does not  
work. i am just wondering why and to which objective there now is this  
collaboration between social network service providers and researchers  
that are keen social network analysts? i am curious as to if this is  
about providing tools that make the information on social networks  
transparent, and if so, for whom? for marketers, scientists,  
politicians etc. more interestingly, who decides what transparency is,  
and how the results are constructed, validated, and made open to  
questioning and contestation. this is obviously not talked about at  
all, there is only a pointer that these researchers are now working  
together with facebook.

on a related note, i am concerned with such studies and their  
construction of statistical identity or truth. first of all, what  
counts as race in itself can be very controversial, as was evident  
from the many points raised in the different emails. quantification of  
race, as necessary as sometimes that may be for equal opportunity  
actions or political projects, also re-produces what race is. so, it  
seems what we have a mix of three things: (1) a construction of race  
(2) a construction of names and racial order, and (3) social networks  
and its user "diversity". the result is a study that melts these three  
elements together. i am just wondering, how such a complex statistical  
and social matter becomes so smooth that it can be fit onto one page  
with no open questions and what are ways of questioning the study. i  
would appreciate any further comments you may have in the direction.

and, once again,  thank you for all the others, especially with  
respect to references on race and class on social networks, the topic,  
regardless of facebooks ability to water it down to some mind turning  
statistics, remains complex and interesting.

s.

***Wherefore art thou r3579x?: anonymized social networks, hidden  
patterns, and structural steganography. Lars Backstrom, Cynthia Dwork,  
Jon M. Kleinberg. WWW 2007


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm