<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    Seems to me much will depend on the beliefs and attitudes of the
    authorities as opposed to the experts. Looking for suspicious
    patterns is traditional in policing and this is just the same --
    except powered by a turbo-charged mechanism. If the cops start to
    assume that the associations the software predicts are pretty well
    infallible, there will be no escape for the innocent and therefore
    this will be Minority Report (1984, The First Circle, etc.)<br>
    <br>
    Fascinating thread (nearly missed because of a most misleading
    subject line)<br>
    <br>
    Cheers -- Mike<br>
    <br>
     <br>
    <div class="moz-cite-prefix">On 23/04/2014 09:44, M.E.Sciubba wrote:<br>
    </div>
    <blockquote
cite="mid:CACTqU0HUdg6_jic111Nhi=yQkzaFK+fUB6RevA-58MC5Zghyxw@mail.gmail.com"
      type="cite">
      <div dir="ltr">Beyond the legal aspect about the prediction of
        human actions and hence loss of basic civil rights, I think the
        key point here is the "cultural" variation (if any) of the
        values attached to what somebody tweets/sends through social
        media. It is a sociological analysis of the interaction between
        people(s) lives and (possible) threatening activities that is
        lacking, if I got it right...
        <div>
          <br>
        </div>
        <div>Very interesting subject indeed, if I could only apply as a
          telecommuter post-doc ;)</div>
        <div><br>
        </div>
        <div>Cheers,</div>
        <div><br>
        </div>
        <div>Eleonora</div>
        <div><br>
        </div>
        <div>____________________________________________________________________<br>
          <div><br>
          </div>
          <div>
            <div id="divRpF498918"
              style="color:rgb(0,0,0);font-family:'Times New
              Roman';font-size:16px;direction:ltr"> </div>
            <div
              style="color:rgb(0,0,0);font-family:Tahoma;font-size:13px">
              <div>Dr. Maria Eleonora Sciubba</div>
              <div>
                <div>Research Associate - Interactional Linguisuistics</div>
                <div>Archivio di LInguA Spontanea</div>
                <div>
                  <div><span
                      style="font-family:Arial,Helvetica,sans-serif;font-size:12px">tel.
                      +32 16 3 24795</span></div>
                  <div><span
                      style="font-family:Arial,Helvetica,sans-serif;font-size:12px">cell
                      +32 468177656</span></div>
                </div>
                <div><br>
                </div>
                <div>
                  <p class="MsoNormal"
style="color:rgb(34,34,34);font-size:12.727272033691406px;font-family:arial,sans-serif"><span
                      lang="EN-US">KU Leuven – Faculty of Arts</span></p>
                  <p class="MsoNormal"
style="color:rgb(34,34,34);font-size:12.727272033691406px;font-family:arial,sans-serif"><span
                      lang="EN-US">Department of French, Italian and
                      Comparative Linguistics</span></p>
                  <p class="MsoNormal"
style="color:rgb(34,34,34);font-size:12.727272033691406px;font-family:arial,sans-serif"><span
                      lang="NL-BE">Blijde-Inkomststraat 21, PO BOX 3308</span></p>
                  <p class="MsoNormal"
style="color:rgb(34,34,34);font-size:12.727272033691406px;font-family:arial,sans-serif"><span
                      lang="NL-BE">B - 3000 Leuven</span></p>
                  <p class="MsoNormal"><span lang="NL-BE"><font
                        color="#222222" face="arial, sans-serif"><a
                          moz-do-not-send="true"
                          href="http://www.kuleuven.be/wieiswie/nl/person/00088846">http://www.kuleuven.be/wieiswie/nl/person/00088846</a></font></span></p>
                </div>
              </div>
            </div>
            <div
              style="color:rgb(0,0,0);font-family:Tahoma;font-size:13px">
              <p class="MsoNormal"><span lang="NL-BE"><font
                    color="#222222" face="arial, sans-serif"><br>
                  </font></span></p>
              <p class="MsoNormal"><span lang="NL-BE"></span></p>
              <hr tabindex="-1" style="font-family:'Times New
                Roman';font-size:16px">
              <div id="divRpF498918" style="font-family:'Times New
                Roman';font-size:16px;direction:ltr"><span
                  style="font-size:13px;font-family:'Segoe
                  UI',Helvetica,Arial,sans-serif">Disclaimer: </span><a
                  moz-do-not-send="true"
href="https://owa.groupware.kuleuven.be/owa/redir.aspx?C=TdLzfiA_2UiPWm_Hku2lb346uUpNcdAINaOnchkCiVs0F2pBz-iwkuvyJDCbnn7U93SVwLpxYn0.&URL=http%3a%2f%2fwww.kuleuven.be%2fcwis%2femail_disclaimer.htm"
                  target="_blank"
                  style="font-size:13px;font-family:'Segoe
                  UI',Helvetica,Arial,sans-serif">http://www.kuleuven.be/cwis/email_disclaimer.htm</a></div>
              <div style="font-family:'Times New Roman';font-size:16px"><i
style="color:rgb(51,204,0);font-family:Tahoma,sans-serif;font-size:medium"><br>
                </i></div>
              <div style="font-family:'Times New Roman';font-size:16px">
                <i
style="color:rgb(51,204,0);font-family:Tahoma,sans-serif;font-size:medium">Be
                  green. Keep it on the screen</i></div>
            </div>
          </div>
        </div>
      </div>
      <div class="gmail_extra"><br>
        <br>
        <div class="gmail_quote">2014-04-23 5:57 GMT+02:00 Alexander Yeh
          <span dir="ltr"><<a moz-do-not-send="true"
              href="mailto:asy@mitre.org" target="_blank">asy@mitre.org</a>></span>:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div class="">Zoltan Boka wrote:<br>
              <blockquote class="gmail_quote" style="margin:0 0 0
                .8ex;border-left:1px #ccc solid;padding-left:1ex">
                Predictions are only as good as the initial data they're
                based on. In<br>
                this case that data could be incomplete or limited or
                biased (for<br>
                instance, lets say that one data point is the number of
                arrests made on<br>
                street x: lets also say that its police policy to harass
                and arrest<br>
                people living on street x- you can see how this can
                infect the process.)<br>
                <br>
                But even if it can be done without bias and with a high
                accuracy, the<br>
                question remains- are some people literally destined to
                commit crime and<br>
                if so can that destiny be altered short of preemptively
                arresting them?<br>
              </blockquote>
              <br>
            </div>
            If you have a guess at a who is likely, you could
            concentrate undercover police to where (and when) that
            person tends to go and also to follow that person around.
            With enough evidence, you could also get search warrants and
            warrants to intercept their phone, etc.<br>
            <br>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">
              <div class="">
                <br>
                Sent from my iPhone<br>
                <br>
                On Apr 22, 2014, at 14:39, Adam Kilgarriff <<a
                  moz-do-not-send="true"
                  href="mailto:adam@lexmasterclass.com" target="_blank">adam@lexmasterclass.com</a><br>
              </div>
              <mailto:<a moz-do-not-send="true"
                href="mailto:adam@lexmasterclass.com" target="_blank">adam@lexmasterclass.com</a>>>
              wrote:<br>
              <br>
              <blockquote class="gmail_quote" style="margin:0 0 0
                .8ex;border-left:1px #ccc solid;padding-left:1ex">
                <div class="">
                  If a clever system can predict who is going to predict
                  a crime - with<br>
                  good, but far from 100%, accuracy, is the use it<br>
                     a) rational policing practice<br>
                     b) discriminatory<br>
                  to use that information?<br>
                  <br>
                  Seems to me, it's both.<br>
                  <br>
                  Marek says<br>
                  > But there are definitely many ways to abuse this
                  technology as well.<br>
                  <br>
                  I don't feel abuse is the main issue.  Any use of it
                  is<br>
                  discriminatory.   Should we trade off? Tough question.<br>
                  <br>
                  Adam<br>
                  <br>
                  <br>
                  On 22 April 2014 11:34, Marek Rei <<a
                    moz-do-not-send="true"
                    href="mailto:marek.rei@gmail.com" target="_blank">marek.rei@gmail.com</a><br>
                </div>
                <div class="">
                  <mailto:<a moz-do-not-send="true"
                    href="mailto:marek.rei@gmail.com" target="_blank">marek.rei@gmail.com</a>>>
                  wrote:<br>
                  <br>
                      Here's an interesting article about how Chicago
                  police is already<br>
                      applying such technology (in somewhat troubling
                  ways):<br>
                  <br>
                      <a moz-do-not-send="true"
href="http://www.theverge.com/2014/2/19/5419854/the-minority-report-this-computer-predicts-crime-but-is-it-racist"
                    target="_blank">http://www.theverge.com/2014/2/19/5419854/the-minority-report-this-computer-predicts-crime-but-is-it-racist</a><br>
                  <br>
                      I wouldn't say crime prediction technology by
                  itself is evil, it's<br>
                      more a question of how it's used. For example, I
                  wouldn't have a<br>
                      problem with a system that can prioritise a large
                  list of likely<br>
                      suspects after a crime has been committed, or is
                  able to flag a<br>
                      social media message calling for a hate crime. But
                  there are<br>
                      definitely many ways to abuse this technology as
                  well.<br>
                  <br>
                      Marek<br>
                  <br>
                  <br>
                  <br>
                      On Tue, Apr 22, 2014 at 10:55 AM, Christian
                  Pietsch<br>
                </div>
                    <<a moz-do-not-send="true"
                  href="mailto:chr.pietsch@googlemail.com"
                  target="_blank">chr.pietsch@googlemail.com</a>
                <mailto:<a moz-do-not-send="true"
                  href="mailto:chr.pietsch@googlemail.com"
                  target="_blank">chr.pietsch@googlemail.com</a>>>
                <div>
                  <div class="h5"><br>
                        wrote:<br>
                    <br>
                            Hi Matthew,<br>
                    <br>
                            so you want to build a heuristic precrime
                    detector based on<br>
                            routine<br>
                            activities observed on social networks. Does
                    that mean that<br>
                            if, say, I<br>
                            tend to update my status at the same time as
                    some terrorist in<br>
                            your<br>
                            training set, your software will label me as
                    a likely<br>
                            terrorist and<br>
                            put me on a no-fly list? Will I get a chance
                    to prove my<br>
                            innocence?<br>
                    <br>
                            When you have some spare time, try to watch
                    Minority Report.<br>
                            Or did<br>
                            this movie inspire your project? Then you
                    must have<br>
                            misunderstood its<br>
                            message.<br>
                    <br>
                            Your suspect<br>
                            Christian<br>
                    <br>
                    <br>
                            On Mon, Apr 21, 2014 at 11:34:11AM -0400,
                    Matthew Gerber wrote:<br>
                            > Hello,<br>
                            ><br>
                            > A new research position has opened
                    within our lab, and we<br>
                            are seeking M.S.,<br>
                            > Ph.D., and post-doc researchers.<br>
                            ><br>
                            > One-sentence summary:  We are mining
                    social media for<br>
                            indicators of<br>
                            > individual routine activities for the
                    purpose of improved<br>
                            crime prediction.<br>
                            ><br>
                            > Longer summary: This project focuses on
                    the spatiotemporal<br>
                            prediction of<br>
                            > localized attacks carried out against
                    individuals in urban<br>
                            areas. We view<br>
                            > an attack as the outcome of a point
                    process governed by the<br>
                            interaction of<br>
                            > attackers, targets, and the physical
                    environment. Our<br>
                            ultimate goal is to<br>
                            > predict future outcomes of this process
                    in order to increase<br>
                            the security<br>
                            > of human populations and U.S. assets
                    and interests. However,<br>
                            achieving this<br>
                            > goal requires a deeper understanding of
                    how attack outcomes<br>
                            correlate with<br>
                            > the routine activities of individuals
                    in an area. The<br>
                            proposed research<br>
                            > will generate this understanding and in
                    doing so will answer<br>
                            questions such<br>
                            > as the following: What are the
                    dimensions along which<br>
                            individuals’<br>
                            > activities should be quantified for the
                    purpose of attack<br>
                            prediction? How<br>
                            > can measurements along these dimensions
                    be taken<br>
                            automatically and with<br>
                            > minimal expense (e.g., via social
                    media)? What are the<br>
                            implications of such<br>
                            > measurements for attack prediction
                    performance? Subsuming<br>
                            these questions<br>
                            > is the issue of geographic variation:
                    do our answers change<br>
                            when moving<br>
                            > from a major U.S. city to a major U.K.
                    city? There has been<br>
                            plenty of<br>
                            > previous work on spatiotemporal attack
                    prediction (see our<br>
                            Asymmetric<br>
                            ><br>
                            Threat<<a moz-do-not-send="true"
href="http://ptl.sys.virginia.edu/ptl/projects/asymmetric-threat-prediction"
                      target="_blank">http://ptl.sys.virginia.edu/ptl/projects/asymmetric-threat-prediction</a>>project);<br>
                            > however, these basic questions remain
                    unanswered, leaving a<br>
                            > substantial gap in our understanding of
                    attack processes and<br>
                            their<br>
                            > relationships with individuals’ routine
                    activities.<br>
                            ><br>
                            > More information can be found<br>
                            ><br>
                            here<<a moz-do-not-send="true"
href="http://ptl.sys.virginia.edu/ptl/projects/routine-activities-analysis-for-crime-prediction"
                      target="_blank">http://ptl.sys.virginia.edu/ptl/projects/routine-activities-analysis-for-crime-prediction</a>><br>
                            > .<br>
                            ><br>
                            > Sincerely,<br>
                            ><br>
                            > Matthew S. Gerber, Ph.D.<br>
                            > Research Assistant Professor<br>
                            > Department of Systems and Information
                    Engineering<br>
                            > University of Virginia<br>
                    <br>
                            > _______________________________________________<br>
                            > UNSUBSCRIBE from this page:<br>
                            <a moz-do-not-send="true"
                      href="http://mailman.uib.no/options/corpora"
                      target="_blank">http://mailman.uib.no/options/corpora</a><br>
                            > Corpora mailing list<br>
                  </div>
                </div>
                        > <a moz-do-not-send="true"
                  href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a>
                <mailto:<a moz-do-not-send="true"
                  href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a>>
                <div class=""><br>
                          > <a moz-do-not-send="true"
                    href="http://mailman.uib.no/listinfo/corpora"
                    target="_blank">http://mailman.uib.no/listinfo/corpora</a><br>
                  <br>
                  <br>
                          _______________________________________________<br>
                          UNSUBSCRIBE from this page: <a
                    moz-do-not-send="true"
                    href="http://mailman.uib.no/options/corpora"
                    target="_blank">http://mailman.uib.no/options/corpora</a><br>
                          Corpora mailing list<br>
                </div>
                        <a moz-do-not-send="true"
                  href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a>
                <mailto:<a moz-do-not-send="true"
                  href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a>>
                <div class=""><br>
                          <a moz-do-not-send="true"
                    href="http://mailman.uib.no/listinfo/corpora"
                    target="_blank">http://mailman.uib.no/listinfo/corpora</a><br>
                  <br>
                  <br>
                  <br>
                      _______________________________________________<br>
                      UNSUBSCRIBE from this page: <a
                    moz-do-not-send="true"
                    href="http://mailman.uib.no/options/corpora"
                    target="_blank">http://mailman.uib.no/options/corpora</a><br>
                      Corpora mailing list<br>
                </div>
                    <a moz-do-not-send="true"
                  href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a>
                <mailto:<a moz-do-not-send="true"
                  href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a>><br>
                    <a moz-do-not-send="true"
                  href="http://mailman.uib.no/listinfo/corpora"
                  target="_blank">http://mailman.uib.no/listinfo/corpora</a><br>
                <br>
                <br>
                <br>
                <br>
                --<br>
                ========================================<br>
                Adam Kilgarriff <<a moz-do-not-send="true"
                  href="http://www.kilgarriff.co.uk/" target="_blank">http://www.kilgarriff.co.uk/</a>>
                <a moz-do-not-send="true"
                  href="mailto:adam@lexmasterclass.com" target="_blank">adam@lexmasterclass.com</a><br>
                <mailto:<a moz-do-not-send="true"
                  href="mailto:adam@lexmasterclass.com" target="_blank">adam@lexmasterclass.com</a>><br>
                Director Lexical Computing Ltd <<a
                  moz-do-not-send="true"
                  href="http://www.sketchengine.co.uk/" target="_blank">http://www.sketchengine.co.uk/</a>><br>
                Visiting Research Fellow University of Leeds <<a
                  moz-do-not-send="true" href="http://leeds.ac.uk"
                  target="_blank">http://leeds.ac.uk</a>><br>
                /Corpora for all/ with the Sketch Engine <<a
                  moz-do-not-send="true"
                  href="http://www.sketchengine.co.uk" target="_blank">http://www.sketchengine.co.uk</a>><br>
                /DANTE: a lexical database for English <<a
                  moz-do-not-send="true" href="http://www.webdante.com"
                  target="_blank">http://www.webdante.com</a>>/
                <div class=""><br>
                  ========================================<br>
                  _______________________________________________<br>
                  UNSUBSCRIBE from this page: <a moz-do-not-send="true"
                    href="http://mailman.uib.no/options/corpora"
                    target="_blank">http://mailman.uib.no/options/corpora</a><br>
                  Corpora mailing list<br>
                </div>
                <a moz-do-not-send="true" href="mailto:Corpora@uib.no"
                  target="_blank">Corpora@uib.no</a> <mailto:<a
                  moz-do-not-send="true" href="mailto:Corpora@uib.no"
                  target="_blank">Corpora@uib.no</a>><br>
                <a moz-do-not-send="true"
                  href="http://mailman.uib.no/listinfo/corpora"
                  target="_blank">http://mailman.uib.no/listinfo/corpora</a><br>
              </blockquote>
              <div class="">
                <br>
                <br>
                _______________________________________________<br>
                UNSUBSCRIBE from this page: <a moz-do-not-send="true"
                  href="http://mailman.uib.no/options/corpora"
                  target="_blank">http://mailman.uib.no/options/corpora</a><br>
                Corpora mailing list<br>
                <a moz-do-not-send="true" href="mailto:Corpora@uib.no"
                  target="_blank">Corpora@uib.no</a><br>
                <a moz-do-not-send="true"
                  href="http://mailman.uib.no/listinfo/corpora"
                  target="_blank">http://mailman.uib.no/listinfo/corpora</a><br>
                <br>
              </div>
            </blockquote>
            <div class="HOEnZb">
              <div class="h5">
                <br>
                <br>
                <br>
                _______________________________________________<br>
                UNSUBSCRIBE from this page: <a moz-do-not-send="true"
                  href="http://mailman.uib.no/options/corpora"
                  target="_blank">http://mailman.uib.no/options/corpora</a><br>
                Corpora mailing list<br>
                <a moz-do-not-send="true" href="mailto:Corpora@uib.no"
                  target="_blank">Corpora@uib.no</a><br>
                <a moz-do-not-send="true"
                  href="http://mailman.uib.no/listinfo/corpora"
                  target="_blank">http://mailman.uib.no/listinfo/corpora</a><br>
              </div>
            </div>
          </blockquote>
        </div>
        <br>
        <br clear="all">
        <div><br>
        </div>
        -- <br>
        <i
style="color:rgb(51,204,0);font-family:Tahoma,sans-serif;font-size:medium">Be
          green. Keep it on the screen</i>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
UNSUBSCRIBE from this page: <a class="moz-txt-link-freetext" href="http://mailman.uib.no/options/corpora">http://mailman.uib.no/options/corpora</a>
Corpora mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Corpora@uib.no">Corpora@uib.no</a>
<a class="moz-txt-link-freetext" href="http://mailman.uib.no/listinfo/corpora">http://mailman.uib.no/listinfo/corpora</a>
</pre>
    </blockquote>
    <br>
    <pre class="moz-signature" cols="72">-- 
Mike Scott

***
If you publish research which uses WordSmith, do let me know so I can include it at
<a class="moz-txt-link-freetext" href="http://www.lexically.net/wordsmith/corpus_linguistics_links/papers_using_wordsmith.htm">http://www.lexically.net/wordsmith/corpus_linguistics_links/papers_using_wordsmith.htm</a>
***
University of Aston and Lexical Analysis Software Ltd.
<a class="moz-txt-link-abbreviated" href="mailto:mike.scott@aston.ac.uk">mike.scott@aston.ac.uk</a>
<a class="moz-txt-link-abbreviated" href="http://www.lexically.net">www.lexically.net</a>
</pre>
  </body>
</html>