Font Size: 

In this paper we present an investigation of the emotional content conveyed by words in online conversations captured on Twitter. A multivariate technique applied to co-occurence of words together with Correspondence Analysis is adopted in order to find clusters of meaningful words detecting emotional categories that provide meaning to everyday events. Specifically, given the current historical period, where the European Union has to gain trust in its citizens, a corpus of 155000 tweets selected through the Italian keywords "Europa" and "EU" is analyzed. Results show clearly how the textual content is structured according to the different emotional expressions.