A FINE-GRAINED ANNOTATION OF IRONY IN ITALIAN SOCIAL MEDIA TEXTS

This master thesis concerns the development of a fine-grained annotation scheme for pragmatic phenomena to be exploited within the context of automatic sentiment analysis, but also to shed some light on these linguistic phenomena. In particular, it will be used to annotate the rhetorical device of irony in texts from social media. Our prime area of investigation is, in fact, the microblogging platform Twitter. From previous research the recognition of irony and the identification of pragmatic and linguistic devices that activate it are known as very challenging tasks to be performed by both humans or automatic tools. Our goal, is to create an annotated Italian corpus, through which we will hopefully be able to resolve some issues concerning irony formalization and automatic detection. This thesis collocates in the context of a multilingual project for studying irony and for developing resources to be exploited in training NLP tools for sentiment analysis.