1st Workshop on Opinion Mining and Sentiment Analysis

From Brede Wiki
Jump to: navigation, search
Event (help)
1st Workshop on Opinion Mining and Sentiment Analysis (WOMSA09)
Location: Seville Spain Map
Date & time:


Link: http://sites.google.com/site/womsa09/
Search: DuckDuckGo Google Bing

1st Workshop on Opinion Mining and Sentiment Analysis (WOMSA09) is a one-day workshop on text sentiment analysis that takes place 2009 November 13 in Seville, Spain. Paper submission deadline is 2009 September 4

The workshop takes place at Escula Téchnica Superior de Ingeniería Informática, University of Seville, and it is part of Conference of the Spanish Association for Artificial Intelligence CAEPIA-TTIA 2009.

The editors for the proceedings were José A. Troyano, Fermín Cruz and Víctor J. Díaz.

[edit] Papers

A co-occurrence based personal sense approach to opinion mining  
by Polina Panicheva and others [1]. They use a corpus by Pang, Lee, Vaithyanathan: "Thumbs up? sentiment classification..." that is 1400 short movie reviews, 700 positive and 700 negative. Lower-cased, stop words elimination. The approach here: "Personal sense".
AffectiveSpace: blending common sense and affective knowledge to perform emotive reasoning 
by E. Cambria and others uses blog post from LiveJournal.
EmotiBlog: a fine-grained model fro emotion detection in non-traditional textual genres 
by Ester Boldrini, Alexandra Balahur and others uses EmotiBlog.
Exploring the use of paragraph-level annotations for sentiment analysis of financial blogs 
by Paul Ferguson and others. These authors reported an application that works from several hundreds financial blogs and with 687 labelled document-topic pairs.
Machine learning techniques for automatic opinion detection in non-traditional textual genres
uses EmotiBlog and uses two machine learning tools, support vector machines and multinomial naive Bayes in an implementation by Weka. Among the results found is that the handling of negation raises the performance by 4% to 7%, though the best system reported seems to be one using stemming.
On the incidence of part-of-speech on polarity identification of user-generated-contexts in Spanish
by Rafael E. Banchs and Joan Codina from Barcelona Media Innovation Centre downloaded 25'330 documents from www.ciao.es where there was associated ratings. Most of the ratings where positive, but they constructed a balanced data set. Some of the tools they used were OpenNLP, PAROLE and TreeTagger.
Personal tools