Análise do Desempenho de Métodos de Análise de Sentimento no Contexto do Sarcasmo - Artigo WICSI2017

Sarcasm is a figure language commonly used in Web especially in online social networks. Currently it has been much discussed in the literature, as it has the ability to turn the polarity of a seemingly positive or negative statement into its opposite. In this context, the present study reports the analysis of the three classification algorithms performance when subjected to classification of messages that have a sarcastic content, thus suffering the impact of polarity inversion caused by the presence of sarcasm. The performance of each algorithm is measured by three evaluation metrics commonly used in the literature, they are: precison, recall and F-measure.