[Air-L] Issues with Facebook's "graph search"? (Nicholas John)

Simon Knight sjgknight at gmail.com
Sun Jul 21 15:09:39 PDT 2013

Hi Nicholas
I don't know how it's being sampled, but I did write some thoughts/bullet
points about risks/potentials of such a tool a while ago
-as-an-epistemic-tool/ which included a link to Brian Kelly's post on its
potential for research developments (how universities might use it for
example).  Given the social, and economic factors potentially involved as
graph search develops I'm not sure I'd use it as a reliable/valid indicator
of taste.  Be interested to see if anyone has already done some work on
it/is planning to.
@sjgknight <http://twitter.com/sjgknight> 


Message: 2
Date: Sun, 21 Jul 2013 11:19:17 +0300
From: Nicholas John <n.john at huji.ac.il>
To: air-l at listserv.aoir.org
Subject: [Air-L] Issues with Facebook's "graph search"?
	<CAN=3fvoqFcphRUUJGmB8oRTpGYsw9DH5-MTAy9JUAfE_z10tmw at mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1

Hi all,
I've just been messing about a bit with Facebook's graph search. According
to it, the music liked by people who like Obama is different from the music
liked by people who like Romney. Who'd have thunk it? (Richard
this with Obama, McCain and MySpace using ElFriendo, so he at the very least
already did.) But does anyone here actually know anything concrete about
graph search and its reliability and validity as a measurer of taste? Does
anyone know how Facebook is doing it? Is it sampling friends of Obama, or
actually looking at all of them and then showing me what music they like?
Obviously I'm not about to launch a huge research project based on results
from Facebook's graph search, or even a small one for that matter, but
perhaps we could think a bit about whether it might have any use for us as
researchers at all, Nik

Assistant Professor
Department of Communication,
The Hebrew University of Jerusalem
Tel: +972-54-7906073

More information about the Air-L mailing list