This Article 
   
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
Seeing Stars of Valence and Arousal in Blog Posts
Jan.-March 2013 (vol. 4 no. 1)
pp. 116-123
Georgios Paltoglou, University of Wolverhampton, Wolverhampton
Michael Thelwall, University of Wolverhampton, Wolverhampton
Sentiment analysis is a growing field of research, driven by both commercial applications and academic interest. In this paper, we explore multiclass classification of diary-like blog posts for the sentiment dimensions of valence and arousal, where the aim of the task is to predict the level of valence and arousal of a post on a ordinal five-level scale, from very negative/low to very positive/high, respectively. We show how to map discrete affective states into ordinal scales in these two dimensions, based on the psychological model of Russell's circumplex model of affect and label a previously available corpus with multidimensional, real-valued annotations. Experimental results using regression and one-versus-all approaches of support vector machine classifiers show that although the latter approach provides better exact ordinal class prediction accuracy, regression techniques tend to make smaller scale errors.
Index Terms:
Mood,Sentiment analysis,Data mining,Algorithm design and analysis,Predictive models,sentiment analysis,Mood,Sentiment analysis,Data mining,Algorithm design and analysis,Predictive models,affect detection,Mining methods and algorithms
Citation:
Georgios Paltoglou, Michael Thelwall, "Seeing Stars of Valence and Arousal in Blog Posts," IEEE Transactions on Affective Computing, vol. 4, no. 1, pp. 116-123, Jan.-March 2013, doi:10.1109/T-AFFC.2012.36
Usage of this product signifies your acceptance of the Terms of Use.