Análise de redes sociais aplicada a tweets sobre séries de tv

Abstract

The widespread use of social networks and the various ways they can be accessed changed the way people and organizations interact with each other, generating a vast amount of information. Many organizations have invested in the analysis of such data in order to obtain information about their clients. The social network analysis and sentiment analysis, along with text mining, provide techniques that can structure and facilitate the understanding of information. Twitter is among the most popular social networks, it is a microblog that allows people publish messages, called tweets, expressing their opinions, making it one of the main sources for the analysis of opinions. In this context, this work aims to collect tweets about TV shows and apply a set of techniques of social networks analysis, sentiment analysis and text mining. The metrics most commonly used in the literature for the analysis of social networks were selected and applied, enabling the analysis of the quantity and strength of the relationship between the TV shows, as well as the popularity of each TV show. For the classification of feelings, messages were categorized with a positive, negative or neutral feeling.