Dipartimento di Economia, Management e Metodi Quantitativi 

Università degli Studi di Milano

terrà il seminario:

 

There are several ways to analyze data coming from digitized texts and there are even more ways to analysis big amount of these data.

Analyzing the sentiment expressed on social networks is both about big data and digitized text. But sentiment analysis is not just text mining, i.e. usual data mining applied to text data, because what matters is to capture the meaning of sentences in a corpora of texts whose language evolves in time and space or according to the age, the background of the writers, etc.

We will discuss the most recent techniques of sentiment analysis available in the literature along with some new ideas and present some applications to the analysis of 80 millions of italian tweets to happyness and economics.