[iDC] sentiment geeks and the social graph

Mark Andrejevic markbandrejevic at gmail.com
Tue Oct 13 09:28:17 UTC 2009


there are really some great conversations taking place on this list -- I'm
trying to keep up! But I also thought I'd add some follow-up thoughts
tangentially related to the exploitation discussion. The familiar framing
of submission to various forms of online monitoring in terms of the logic of
exchange (we submit to the collection of information about ourselves in
return for access to "free" goods and services) needs further interrogation:
not just in terms of what information is collected vs. what information
we consciously disclose about ourselves, and not just in terms of
the economic and social relations that structure the "free" exchange, but
also, perhaps, in terms of the split between the forms of
gratification associated with online services and the data gathered about
us. These might be seen, increasingly, as overlapping categories.

I'm thinking here of a constellations of developments associated with
so-called sentiment analysis: the use of the internet as means not just for
gathering information, but for measuring sentiment. For starters, we might
include in this category Mark Zuckerberg's conception of Facebook as a means
of reconstituting the organization of information online in terms of a
"social graph" -- a means of organizing information and facilitating
searches based not on, as Wired magazine puts it, "the cold mathematics" of
a Google search, but on a more "personalized, humanized" algorithm that
draws on our social networks to shape our searches and provide us with
customized results. Alongside this individual use of the social graph is the
goal of enlisting so-called sentiment analysis -- an attempt to gauge
sentiment by sorting through large-scale databses (what Pang and Lee have
called "opinion mining") -- for marketing purposes. Companies like Jodange
and Scout Labs (which I learned about through a NYTimes piece on sentiment
analysis), promise a kind of gestalt reading of the data flow: a means of
seeing the whole without necessarily having to read through all the discreet
data, that is reminsicent of the new spate of attempts to privilege gut
instinct, first impressions, body language, etc. (as outlined, for example
in Gladwell's Blink and represented in a spate of shows about adepts who are
able to beat the machines -- the cold mathematics of the algorithm
-- through their ability to read emotions and gauge impressions -- Lie to
Me, the Mentalist, etc.).

So Jodange, for example describes its goal as: the development of "business
applications that drive tangible value by allowing knowledge workers to
better understand who is influencing their customers, competitors and
marketplace in an environment where information continues to originate from
an exploding number of information sources" and Scout Labs promises to help
clients track social media and "find signals in the noise to help your team
build better products and stronger customer relationships." The site
includes the following (anonymous -- maybe the web is speaking) testimonial:
"Scout Labs provides an intuitive and elegant interface for managing a
wealth of conversations across the web. It makes social media monitoring
dead-simple."

I'm not sure what this adds to the portrayal of interactive applications as
(among other things) a means of gathering information about consumers -- but
I'm intrigued that the goal is not simply demographics, patterns of browsing
or purchasing behavior, and not even the collection of data about indvidual
preferences, but the goal of discerning in the data flow a dominant feeling
tone coalescing around particular products, initiatives, people or
campaigns. At work here is a kind of prosopopeia in the sense in which Zizek
has been using it recently -- the creation of some kind of aggregate
non-subject whose sentiments can be read off the data.

It's hard, when looking at these developments, not to be struck by Patricia
Ticineto Clough's observation that, "this is a dynamic background, a
probablisitc, statistical background which provides an infra-empirical or
infra-temporal sociality, the subject of which is, I want to propose, the
population, technologically or methodologically open to the modulation of
its affective capacities. Sociality as affective background displaces
sociality grasped in terms off structure and individual; affective
modulation and individuation displace subject formation and ideological
interpellation as central to the relation of governance and economy" (from
The New Empiricism: Affect and Sociological Method, European Journal of
Social Theory 2009).



**
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.thing.net/pipermail/idc/attachments/20091013/c98ac275/attachment.htm 


More information about the iDC mailing list