<div dir="ltr">A new dissertation on this topic -- and in fact involving some experimentation specifically using texts on capital punishment -- was completed recently:<br><br><div style="margin-left: 40px;">Stephan Greene, <span style="font-size: 11pt;"><i>Spin: Lexical Semantics, Transitivity, and
the Identification of Implicit Sentiment</i></span>. Unpublished doctoral dissertation, Department of Linguistics, University of Maryland, 2007. <br><a href="http://www.umiacs.umd.edu/~sgreene/SGreeneDissertationFinalDist.pdf">http://www.umiacs.umd.edu/~sgreene/SGreeneDissertationFinalDist.pdf</a><br>
</div><br>A one-page dissertation abstract is at <a href="http://www.umiacs.umd.edu/~sgreene/SGreeneDissertationFinalDist-Abstract.pdf">http://www.umiacs.umd.edu/~sgreene/SGreeneDissertationFinalDist-Abstract.pdf</a>.<br>
<br>More generally, you'll want to see the excellent and very comprehensive monograph recently published by Bo Pang and Lillian Lee: <br><br><div style="margin-left: 40px;"><i><a href="http://www.cs.cornell.edu/home/llee/opinion-mining-sentiment-analysis-survey.html">http://www.cs.cornell.edu/home/llee/opinion-mining-sentiment-analysis-survey.html</a></i><br>
</div><br><br> Philip<br> <a href="mailto:resnik@umd.edu">resnik@umd.edu</a><br><br><br><br><div class="gmail_quote">On Wed, Oct 22, 2008 at 8:44 AM, Anthony Jappy <span dir="ltr"><<a href="mailto:tony@univ-perp.fr">tony@univ-perp.fr</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hi,<br>
I have a colleague who has been manually analyzing texts in order to<br>
determine whether they are positive or negative with respect to their central<br>
theme. For obvious reasons, he would now like to automate the process.<br>
Does anyone know of any way of determining whether the texts in an electronic<br>
corpus are positive or negative with respect to their theme: i.e. given a set<br>
of texts on capital punishment, is there any way of determining reasonably<br>
accurately whether author A's contribution can be seen as more positive than<br>
author B's, or whether a given text is positive or downright negative? My only<br>
experience in this area is with attempts to replicate with students reported<br>
results concerning semantic prosodies, certainly not the most economical way of<br>
going about the problem mentioned above.<br>
There may have been a thread on this problem, but I'm afraid I don't recall<br>
it.<br>
Any bibliographical references would be gratefully accepted.<br>
<br>
TJ<br>
<br>
****************************************************************************<br>
Tony Jappy<br>
<br>
Department of English and North-American Studies,<br>
University of Perpignan-Via Domitia,<br>
66860 Perpignan Cedex,<br>
France<br>
tel : +33 (0)4 68 66 22 76<br>
fax : +33 (0)4 68 66 17 28<br>
e-mail : <a href="mailto:tony@univ-perp.fr">tony@univ-perp.fr</a><br>
<br>
****************************************************************************<br>
<br>
<br>
<br>
<br>
<br>
--<br>
Ce message a été vérifié par MailScanner<br>
pour des virus ou des polluriels et rien de<br>
suspect n'a été trouvé.<br>
CRI UPVD <a href="http://www.univ-perp.fr" target="_blank">http://www.univ-perp.fr</a><br>
<br>
<br>
_______________________________________________<br>
Corpora mailing list<br>
<a href="mailto:Corpora@uib.no">Corpora@uib.no</a><br>
<a href="http://mailman.uib.no/listinfo/corpora" target="_blank">http://mailman.uib.no/listinfo/corpora</a><br>
<br>
</blockquote></div><br></div>