A study on irony within the context of 7x1-PT corpus
Resumo
The increasing use of social networks to express consumer opinions yields a large amount of potentially useful information for organizations to gauge consumer perception of their products. Nevertheless, gauging information by assigning polarities to opinionated text is not a trivial task, especially when dealing with short and ironical text. In this paper, we evaluate the presence of irony at the sentence level within a Portuguese corpus extracted from Twitter.