I was concerned about accessibility issues with Turing tests for C/R
systems (this is related to the CRI protocol me and Eric Dean are working
on). I contacted the relevant person at W3C and posting the response to the
group.
Yakov
Date: Wed, 02 Jul 2003 18:12:29 -0400
To: Yakov Shafranovich <research(_at_)solidmatrix(_dot_)com>
Subject: Re: Fwd: [Asrg] News Article - Turing Tests and the Blind
Yakov,
To clarify what Paul said in the CNET article: the W3C/WAI Protocols and
Formats Working Group is _thinking_ about the issue; and the Web Content
Accessibility Guidelines Working Group (WCAG WG) has the issue on their
issue tracking list. Various informal conversations have taken place
outside of those settings as well. I would be happy to fill you in more
and to connect you with the relevant people. It is part of a general
problem with the intersection of information assurance and
accessible/universal design.
I told Paul that if the WCAG WG comes up with effective _techniques_ to
address this problem, they would likely be included in an updated Working
Draft of the Techniques document that accompanies a future Working Draft
of the Web Content Accessibility Guidelines 2.0. W3C does not anticipate
developing a "standard for Turing tests" -- the techniques layer of our
WAI Guidelines are non-normative. But I do expect that we can contribute
ideas to the development of effective solutions for this problem.
---snip---
At 12:12 PM 7/2/2003 -0400, Yakov Shafranovich wrote:
Hi,
I am a member of the Anti Spam Research Group (ASRG) of the IRTF
(www.irtf.org/asrg/). I recently saw you being quoted in the following
news article regarding developing web accessibility standards for Turing
tests. If it possible, can you point me to the specific working groups
and standards that are working on this issue. Also, if it possible can
you summarize for the our group what exactly is W3C planning on doing
regarding Turing tests.
Yakov
To: asrg(_at_)ietf(_dot_)org
From: Yakov Shafranovich <research(_at_)solidmatrix(_dot_)com>
Subject: [Asrg] News Article - Turing Tests and the Blind
Sender: asrg-admin(_at_)ietf(_dot_)org
List-Id: Anti-Spam Research Group - IRTF <asrg.ietf.org>
List-Archive: <https://www1.ietf.org/pipermail/asrg/>
Date: Wed, 02 Jul 2003 12:03:19 -0400
This News.com article (http://news.com.com/2100-1032_3-1022814.html)
discussed a problem with using visual Turing tests - the blind cannot use it.
---snip---
Efforts to create tests aimed at distinguishing humans from machines go
back decades, with the most famous formulation of the problem posed in
1950 by the English mathematician and World War II "Enigma" code breaker
Alan Turing. Turing's controversial hypothesis was that a machine could
be defined as "intelligent" if a questioner could be fooled into
believing it was a person.
Visual tests in a sense turn that theory on its head, assuming that a
machine is defined by its inability to perform a task that is easy for
most humans to accomplish.
The increasingly popular visual test, and the difficulty of using
current work-arounds, has raised enough hackles among advocates for the
disabled that working groups within the World Wide Web Consortium's
(W3C) Web Accessibility Initiative have begun discussions on how to
standardize an alternative.
Two WAI working groups are hashing out proposals to guide Web sites in
designing blind-friendly bot repellants, and the WAI hopes to address
the issue in the next working draft of its Web Accessibility Guidelines,
Version 2.0, which is due by year's end. So far, published working
drafts of the guidelines are silent on the issue.
----snip----
As Vernon pointed out previously we do not know for sure if these tests
cannot be passed by a machine. Also, nothing can stop spammers from
hiring real people to do the task. The financial difference would be
negligible. This raises the issue of whether the tests are needed in the
first place. C/R systems also fall under this issue.
It would be worthwhile to contact the W3C in order to keep abreast of
any standards they make in the area.
Yakov
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg