[Humanist] 26.76 name for a kind of data

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Sun Jun 10 22:19:31 CEST 2012


                  Humanist Discussion Group, Vol. 26, No. 76.
            Department of Digital Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org

  [1]   From:    Ben Miller <bjmiller at MIT.EDU>                             (58)
        Subject: Re: [Humanist] 26.72 name for a kind of data?

  [2]   From:    Maximilian Schich <maximilian at schich.info>                (47)
        Subject: Re: [Humanist] 26.72 name for a kind of data?

  [3]   From:    Alan Corre <corre at uwm.edu>                                 (2)
        Subject: Vol. 26, No. 72 A kind of data


--[1]------------------------------------------------------------------------
        Date: Sat, 09 Jun 2012 22:18:56 -0400
        From: Ben Miller <bjmiller at MIT.EDU>
        Subject: Re: [Humanist] 26.72 name for a kind of data?
        In-Reply-To: <20120609204825.4436E37185 at woodward.joyent.us>

Hello,

If I understand the question correctly, the general category for most of 
the data types collected by Facebook (and other social media sites) is 
"transactional data."  A recent (2002) invocation of the value of that 
category of data was by John Poindexter.  Yes, that is the same John 
Poindexter made infamous by his enabling role as National Security 
Advisor in Regan's Iran-Contra scandal, who later went on to build a 
data-mining project for DARPA called Total Information Awareness.  One 
could justifiably imagine he learned of the value of this data whilst 
standing over a shredder in a backroom of the Pentagon.  Transactional 
data is what gets created when an entity performs a system operation.  
The term emerged from economic literature in the late 1800s, but only 
infrequently, and didn't become more commonplace until information 
engineering took off in the 1950s-80s.  A purchase with anything from a 
credit card to bitcoins is a transaction; friending or unfriending a 
person on Facebook is a transaction.  Each of these events requires a 
similar set of database operations.

A colleague of Poindexter's on one branch of the TIA project ("Evidence 
Extraction and Link Discovery Program") provides a wonderful example of  
the value of that data here: 
 http://archive.darpa.mil/DARPATech2002/presentations/iao_pdf/speeches/SENATOR.pdf 
Schrems' delineates 84 specific categories of Facebook data here: 
 http://www.europe-v-facebook.org/fb_cat1.pdf 

This category of data, I would argue, constitutes something like an 
on-line biography, but it not itself biodata.  It is the very detailed 
shape of the trail generated by the intersection of agency and system, 
not the data that describes the agent or the system outside of their 
interactions.  Much of what Schrems' access request generated is not 
biographic data, but transactional data.  I would argue there is not a 
good umbrella term to describe the extensive variety of data types we 
generate while we inhabit these virtual communities.

Best,
Ben

--
Ben Miller, Ph.D.
Assistant Professor of English and Communication
Co-Director, Second Century Initiative in New and Emerging Media
Georgia State University

On 6/9/2012 4:48 PM, Humanist Discussion Group wrote:
>          Date: Fri, 08 Jun 2012 19:31:41 +1000
>          From: Willard McCarty<willard.mccarty at mccarty.org.uk>
>          Subject: a name for a kind of data?
>
> A question, as follows, from an MA student at King's:
>
>> I was wondering if you would know the exact name of data that is
>> stored of people that is not immediately accessible to them but
>> stands as a record of their online activities. An example would be
>> the 1200 pages of personal biodata that Facebook user, Max Schrems
>> was sent when he requested for all his personal information, before
>> it was consciously retrieved by Facebook/Max Schrems but just exists
>> as useless/potentially useful information 'floating' in the
>> internet.
> Any ideas? Please reply to wei_jie_mark.tan at kcl.ac.uk as well as to
> Humanist.
>
> Yours,
> WM



--[2]------------------------------------------------------------------------
        Date: Sun, 10 Jun 2012 10:26:10 +0200
        From: Maximilian Schich <maximilian at schich.info>
        Subject: Re: [Humanist] 26.72 name for a kind of data?
        In-Reply-To: <20120609204825.4436E37185 at woodward.joyent.us>

This kind of data is actually not as new as the public attention to it:

I guess the most accurate name is User Activity Data:
cf. http://www.jisc.ac.uk/events/2010/07/businessintelligence.aspx
or 
http://blog.programmableweb.com/2009/04/28/facebook-opens-user-activity-stream-to-developers/

A related concept is User Clickstream Data:
cf. http://en.wikipedia.org/wiki/Clickstream
or http://www.plosone.org/article/info:doi/10.1371/journal.pone.0004803

Another important concept in this circumstance is Reality Mining:
For a brilliant and concise introduction see Sandy Pentland's book on 
Honest Signals:
cf. http://www.amazon.com/Honest-Signals-Shape-World-Bradford/dp/0262162563
or http://www.youtube.com/watch?v=T1iKKAA2FOw

Best, Max

Register now: http://artshumanities.netsci2012.net

-- 
Dr. Maximilian Schich
SOMS, ETH Zurich
http://www.soms.ethz.ch
http://www.schich.info


--[3]------------------------------------------------------------------------
        Date: Sun, 10 Jun 2012 13:41:30 -0500 (CDT)
        From: Alan Corre <corre at uwm.edu>
        Subject: Vol. 26, No. 72 A kind of data
        In-Reply-To: <20120609204825.4436E37185 at woodward.joyent.us>

How about "cyberdebris"?

Alan Corré





More information about the Humanist mailing list