[Humanist] 23.749 meaning by images; Recaptcha

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Tue Apr 6 08:03:26 CEST 2010


                 Humanist Discussion Group, Vol. 23, No. 749.
         Centre for Computing in the Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org

  [1]   From:    Jeff Drouin <jdrouin at gc.cuny.edu>                         (55)
        Subject: Re: [Humanist] 23.747 transposition of meaning by images?

  [2]   From:    Peter Organisciak <organisc at ualberta.ca>                  (27)
        Subject: Re: [Humanist] 23.745 Recaptcha?


--[1]------------------------------------------------------------------------
        Date: Mon, 5 Apr 2010 09:28:59 -0400
        From: Jeff Drouin <jdrouin at gc.cuny.edu>
        Subject: Re: [Humanist] 23.747 transposition of meaning by images?
        In-Reply-To: <20100405070641.99E7A4F232 at woodward.joyent.us>

Dear Susan,

I'm not sure if this is directly relevant to your student's query, but one of 
my projects, the Ecclesiastical Proust Archive, uses images to interpret the 
church motif in A la recherche du temps perdu. It consists of a searchable 
database of all church-related passages, paired with photographs showing the 
original churches or, where there is none, some architectural or other 
feature that captures the emotive or narrative spirit of the passage.

http://proustarchive.org

The project exists as a proof of concept and will be drastically changed and 
expanded over the coming year.

The Rationale page provides a window into my thinking behind representational 
vs. interpretive images. However, there is much more to it, so your student 
should feel free to contact me directly at jsdrouin at gmail dot com.

Best,

Jeff

|| On Monday 05 April 2010 03:06:41 Humanist Discussion Group wrote:
                 Humanist Discussion Group, Vol. 23, No. 747.
         Centre for Computing in the Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org

        Date: Sun, 4 Apr 2010 18:31:16 -0400
        From: Susan Brown <sbrown at uoguelph.ca>
        Subject: image transposition on the web

Does anyone know of any useful work (formal or informal) about the
transposition of meaning in digital media through the
recontextualization of a digital image? I have a graduate student who
wants to consider the movement of meaning both in the works of John
Cayley and on websites such as Tumblr and Ffffound, which reproduce
and reframe already existent images, and he's been having difficulty
finding material with which to frame his argument. Any suggestions of
analyses that might be helpful, including work on mashups or phenomena
such as reblogging/retweeting, would be very welcome, as would
considerations of the reproduction of images on the web. He's
interested in considering the transposition of meaning as the site of
meaning itself.

Many thanks,
Susan
_________________________________________________________________

Susan Brown
Director, Orlando Project; Project Leader, Canadian Writing Research
Collaboratory
Professor                                    				Visiting Professor
School of English and Theatre Studies   	English and Film Studies
University of Guelph                    			University of Alberta
Guelph, Ontario N1G 2W1 Canada          	Edmonton, Alberta T6G 2E5

519-824-4120 x53266 (office)            		780-862-0155
519-766-0844 (fax)

sbrown at uoguelph.ca                      			susan.brown at ualberta.ca
http://orlando.cambridge.org  http://www.ualberta.ca/ORLANDO



--[2]------------------------------------------------------------------------
        Date: Mon, 5 Apr 2010 13:11:57 -0600
        From: Peter Organisciak <organisc at ualberta.ca>
        Subject: Re: [Humanist] 23.745 Recaptcha?
        In-Reply-To: <20100405070641.99E7A4F232 at woodward.joyent.us>


>
> I'd like to find out about the use of ReCaptcha in humanities projects of
> various kinds or sizes. Are there any? Does it help with cleaning up OCR,
> especially with texts before 1800?
>
> Martin Mueller
>

Martin,

Though ReCaptcha is free to use, it is only within its capacity as a CAPTCHA
— which is to say as a layar between users and a website to keep bots out.
In implementing it, you can't control the texts being digitized. Also, since
Google acquired them, I don't know if asking nicely would be effective. I
came across a fellow that wanted to licence the ESP Game (an encoding
project with the same PI as ReCaptcha) but was unable to because Google had
also licensed that project. I can't confirm this, so take it as you will.

As for your question, reCaptcha reported 99.1% accuracy in this Science
article: http://www.sciencemag.org/cgi/content/abstract/321/5895/1465. If I
recall, it used to be run on Internet Archive texts, and is now plowing
through the New York Times archives: much more contemporary than that what
you're asking about. However, if you are interested in building something
similar, the fact that words are independently confirmed by multiple users
cannot be understated, and I'd think it's reasonable to think that it would
at least make worthwhile dent in texts before 1800. However, a point to
remember if trying it yourself is that reCaptcha is constantly evolving to
keep ahead of spammers, though a little fish implementation likely wouldn't
have to worry about that.

Peter





More information about the Humanist mailing list