Twitter for Sentiment Analysis: When Language Resources are Not Available

Affective lexicons are a useful tool for emotion studies as well as for opinion mining and sentiment analysis. Such lexicons contain lists of words annotated with their emotional assessments. There exist a number of affective lexicons for English, Spanish, German and other languages. However, only a few of such resources are available for French. A lot of human efforts are needed to build and extend an affective lexicon. In our research, we propose to use Twitter, the most popular microblogging platform nowadays, to collect a dataset of emotional texts in French. Using the collected dataset, we estimated affective norms of words to construct an affective lexicon, which we use for polarity classification of video game reviews. Experimental results show that our method performs comparably to classic supervised learning methods.

[1]  Jaime Redondo,et al.  The Spanish adaptation of ANEW (Affective Norms for English Words) , 2007, Behavior research methods.

[2]  Chih-Jen Lin,et al.  LIBLINEAR: A Library for Large Linear Classification , 2008, J. Mach. Learn. Res..

[3]  Patrick Paroubek,et al.  Twitter as a Corpus for Sentiment Analysis and Opinion Mining , 2010, LREC.

[4]  Bo Pang,et al.  Thumbs up? Sentiment Classification using Machine Learning Techniques , 2002, EMNLP.

[5]  Mike Thelwall,et al.  A Study of Information Retrieval Weighting Schemes for Sentiment Analysis , 2010, ACL.

[6]  Jonathon Read,et al.  Using Emoticons to Reduce Dependency in Machine Learning Techniques for Sentiment Classification , 2005, ACL.

[7]  Maite Taboada,et al.  Not All Words Are Created Equal: Extracting Semantic Orientation as a Function of Adjective Relevance , 2007, Australian Conference on Artificial Intelligence.

[8]  M. Kendall.  A NEW MEASURE OF RANK CORRELATION , 1938 .

[9]  Sung-Hyon Myaeng,et al.  Determining Mood for a Blog by Combining Multiple Sources of Evidence , 2007 .

[10]  Patrick Paroubek,et al.  Annotations for Opinion Mining Evaluation in the Industrial Context of the DOXA project , 2010, LREC.

[11]  Carlo Strapparava,et al.  WordNet Affect: an Affective Extension of WordNet , 2004, LREC.

[12]  J. S. Maritz.  Distribution-Free Statistical Methods , 1981 .

[13]  Yvette Yannick Mathieu,et al.  A Computational Semantic Lexicon of French Verbs of Emotion , 2006, Computing Attitude and Affect in Text.

[14]  Ling Lin,et al.  Leveraging Webpage Classification for Data Object Recognition , 2007 .

[15]  M. Bradley,et al.  Affective Norms for English Words (ANEW): Instruction Manual and Affective Ratings , 1999 .

[16]  A. Jacobs,et al.  The Berlin Affective Word List Reloaded (BAWL-R) , 2009, Behavior research methods.

[17]  Gérard Dray,et al.  Détection d'opinion : Apprenons les bons adjectifs ! , 2008 .