[Corpora-List] Call for Papers: Mining User-GeneratedContent for Security - MINUCS 2009

[Yannick Versley]

I think there's several aspects that we should try to separate out:

* one is the aspect that NLP and IR tools are used, and will be used,
  for practical purposes. By individuals, companies, and governments.
  This is also one of the reasons why computational linguists enjoy a
  bit more funding than, say, classical Greek historians.
  I also think there's a merit to organize a workshop specifically for
  *real world applications* of IR/NLP techniques instead of the pre-
  fabricated toy problems that CL usually works on. (Hint: They're
  "toy" problems because the developer can hide behind Dijkstra's
  wall and pretend that all that counts is evaluation figures on his
  testing data, without questioning the representativity of the testing
  data, the method of evaluation, or the sensibility of the task as a
  whole. That doesn't mean the problems can be challenging and
  difficult, just that there is an additional step to make if you want
  to apply these standardized problems to anything in the real world,
  and typically you would choose different tradeoffs for recall, computing
  time, and necessary human intervention than if you do CL research).
* The second point is whether these techniques are needed. They are.
  I think the workshop organizers put some thought into the list of
  applications that they presented in the CfP.
  As an additional application, consider the prevention and also the
  aftermath of genocides (which still happen in our nice cozy world,
  possibly ignored by the media at large).
From:
http://daily.swarthmore.edu/2005/2/17/un-advisor-on-genocide-commends-swat-students-illustrates-challenges-of-international-action/

> In finding information about potential and ongoing genocide he said he had
> to both sort through the diverse, voluminous system of information retrieval
> the UN currently has and find information outside the system in order to
> alert the international community to unknown genocidal situations. To
> maintain the trust of the UN and his power to influence action, Mendez
> stressed the fact that he had to closely examine the veracity of information
> and also give reasonable, creative suggestions that are both within the
> limits of what the Security Council will agree to enact and which will not
> make the situation worse.
>
With journalism increasingly being replaced by cute kittens and
drivel-mongering due to the decreasing number of people willing to pay for a
newspaper when they
can just switch on the TV or their internets, retrieval of information, but
also methods
that helps judge its credibility ("I read it on the internet"), will be
essential in maintaining the ability of people to have a critical, informed
perspective on the world.
* Now to the last point: I think that the "Mining ... for Security" is very
poorly chosen.
  It immediately brings up the reading of Security as 'anything that my
friendly funding agency needs to do to save the day', or possibly, anything
that another agency in a non-democratic regime (who will also buy our
system) needs to do.
It is a shame that security, as the security and well-being of citizens, is
no longer the preferred reading for constructions in this kind, but that's
how it is.
And if you, as a workshop organizer, think it's sufficiently ambiguous that
you need to add a disclaimer that it's not for discussion of nilly-willy
ethical issues: well,
maybe you can *also* see that your workshop title is poorly chosen if you
think
it brings on thoughts about ethical issues. If you wrote
"Mining user-generated content for real-world purposes", no one would think
about
poorly-legitimated three letter agencies and opressive regimes in need of
censorship and detection of dissidents.
To everyone else: if you think that your research cannot be used for
practical purposes (good or bad), re-read points one and two.

Just my two cents, and sorry for the long post.

Best wishes,
Yannick Versley

I think you missed the point. To quote:
>
> " Submissions that focus on legal questions stemming from
> snooping, spying, privacy infringement or violation, etc., will not
> be considered relevant to the Theme of the Workshop, and the
> Committee will not be able to review them."
>
> That is the outrageous statement that lead to the distress.  It is
> a matter of fact that the technologies being reviewed here are
> being used for "snooping, spying, privacy infringement or violation,
> etc." In case you haven't been reading the newspapers, there
> was a recent election in Iran, wherein an earlier version of this
> kind of technology was used to suppress news and to find and
> jail dissidents, some of whom were apparently beaten to death.
>
> These technologies can be, and are being used in anti-democratic,
> anti-persona-freedom kinds of ways. Researchers such as you,
> who seem to be unaware of this, should become aware.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20090814/32950dbd/attachment.htm>
-------------- next part --------------
_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora