<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"><head><meta http-equiv=Content-Type content="text/html; charset=utf-8"><meta name=Generator content="Microsoft Word 14 (filtered medium)"><style><!--
/* Font Definitions */
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
@font-face
{font-family:Tahoma;
panose-1:2 11 6 4 3 5 4 4 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0cm;
margin-bottom:.0001pt;
font-size:12.0pt;
font-family:"Times New Roman","serif";}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:purple;
text-decoration:underline;}
span.hoenzb
{mso-style-name:hoenzb;}
span.EmailStyle18
{mso-style-type:personal-reply;
font-family:"Calibri","sans-serif";
color:#1F497D;}
.MsoChpDefault
{mso-style-type:export-only;
font-family:"Calibri","sans-serif";
mso-fareast-language:EN-US;}
@page WordSection1
{size:612.0pt 792.0pt;
margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]--></head><body lang=EN-GB link=blue vlink=purple><div class=WordSection1><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Hello Eugenio,<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>I fear that your plans will result in a corpus that is moderately useful. <o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>If you select and annotate only the positive and negative tweets, you can use the result to learn how to distinguish the positive from the negative, but you will not have any data to learn how to distinguish these subjective tweets from the neutral ones. This latter group is important to recognise as it presumably is the majority class. I cannot imagine how the positive-negative distinction is useful if you do not also distinguish the neutral and the subjective cases. <o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Another issue is that – using your planned method – you will end up with a non-realistic set of positive and negative tweets, as they will only be those where sentiment is expressed lexically. Any experimental results based on that biased corpus will not be representative for real-life texts.<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>It is much more work to annotate all samples from a random selection of texts or snippets, but I believe that this is what you will eventually need to do.<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Greetings to Jaén,<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Ralf<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>European Commission – Joint Research Centre (JRC)<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Ispra, Italy<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'><o:p> </o:p></span></p><p class=MsoNormal><b><span lang=EN-US style='font-size:10.0pt;font-family:"Tahoma","sans-serif"'>From:</span></b><span lang=EN-US style='font-size:10.0pt;font-family:"Tahoma","sans-serif"'> corpora-bounces@uib.no [mailto:corpora-bounces@uib.no] <b>On Behalf Of </b>Eugenio Martínez Cámara<br><b>Sent:</b> 17 December 2011 20:09<br><b>To:</b> Diana Maynard<br><b>Cc:</b> corpora@uib.no<br><b>Subject:</b> Re: [Corpora-List] Spanish Twitter Lexicon<o:p></o:p></span></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>Thanks Diana for your response and your paper.<o:p></o:p></p><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>I explain you what I want to do. I have done several experiments with tweets in Spanish following a machine learning approach, but the problem is I don't have a corpus with a reliable labelling, so I want to build a corpus with a manual labelling. So I've downloaded a set of politic tweets during the last Spanish elections. For the manual labelling process, I want to automatically delete those tweets that aren't opinions. So I'm looking for a Spanish or English word list of opinion words. If a tweet doesn't contain any opinion word I consider that it isn't an opinion tweet. I know that a person can express a politic opinion without using any typical opinion word, but it is a simple heuristic to reduce the set of tweets to be manually labelling.<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>Regards.<o:p></o:p></p></div><div><p class=MsoNormal><br clear=all>Eugenio Martínez Cámara.<o:p></o:p></p><div><p class=MsoNormal>Grupo de Investigación SINAI.<o:p></o:p></p></div><div><p class=MsoNormal>Departamento de Informática.<o:p></o:p></p><div><p class=MsoNormal>Universidad de Jaén.<o:p></o:p></p></div><div><p class=MsoNormal style='margin-bottom:12.0pt'>emcamara at ujaen dot es<o:p></o:p></p></div></div><p class=MsoNormal style='margin-bottom:12.0pt'><br><br><o:p></o:p></p><div><p class=MsoNormal>El 17 de diciembre de 2011 19:40, Diana Maynard <<a href="mailto:d.maynard@dcs.shef.ac.uk">d.maynard@dcs.shef.ac.uk</a>> escribió:<o:p></o:p></p><p class=MsoNormal>Hi Eugenio<br>Are you asking for some gazetteer list of opinionated words to determine whether a tweet is opinionated or not? Or are you asking for some method which uses bag-of-words (matching against such a list) in order to compare your tools with.<br>If the former, obviously you want to be very careful about using such an approach on its own, because there are lots of words which can convey an opinion or not, depending how they are used.<br><br>I am also working on opinion mining from tweets, for English and German, on political tweets amongst other things. You can see my paper about this for English here:<br><br>D. Maynard and A. Funk. Automatic detection of political opinions in tweets. In Proceedings of MSM 2011: Making Sense of Microposts. Workshop at 8th Extended Semantic Web Conference (ESWC 2011). Heraklion, Greece. June 2011.<br><a href="http://gate.ac.uk/sale/eswc11/opinion-mining.pdf" target="_blank">http://gate.ac.uk/sale/eswc11/opinion-mining.pdf</a><br><br>There is also an extended version currently in press.<br>Regards<span style='color:#888888'><br><span class=hoenzb>Diana</span></span><o:p></o:p></p><div><div><p class=MsoNormal><br><br><br>On 17/12/2011 16:05, Eugenio Martínez Cámara wrote:<o:p></o:p></p></div></div><blockquote style='border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm'><div><div><p class=MsoNormal style='margin-bottom:12.0pt'>Hi All,<br><br>Currently I'm working in Sentiment Analysis on Twitter. I have done<br>several experiments with Spanish Twitter corpus following the Go et al.<br>(2009) noisy labels technique, but I want to build a gold standard. So,<br>I downloaded a corpus of Spanish tweets in the politic domain. At first,<br>I want to erase all non-opinion tweets, so I'm going to delete all<br>tweets that not contain any opinion word. So, do you know any Spanish<br>opinion bag-of-words (positive/negative)? do you know any English<br>opinion bag-of-words (positive/negative)?<br><br>Thanks.<br><br><br>Eugenio Martínez Cámara.<br>SINAI Research Group<br>Computer Science Department<br>University of Jaén<br>emcamara at ujaen dot es<br><br><br><br><o:p></o:p></p></div></div><div><p class=MsoNormal>_______________________________________________<br>UNSUBSCRIBE from this page: <a href="http://mailman.uib.no/options/corpora" target="_blank">http://mailman.uib.no/options/corpora</a><br>Corpora mailing list<br><a href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a><br><a href="http://mailman.uib.no/listinfo/corpora" target="_blank">http://mailman.uib.no/listinfo/corpora</a><o:p></o:p></p></div></blockquote><p class=MsoNormal><o:p> </o:p></p></div><p class=MsoNormal><o:p> </o:p></p></div></div></body></html>