Data mining of social manifestations in Twitter: Analysis and aspects of the social movement “Bela, recatada e do lar” (Beautiful, demure and housewife)

Abstract

In recent years, the Online Social Networks (OSN) enabled the growth of many social movements in digital media, because they allow messages to be posted and shared instantly. In the political scope, the OSN are a means by which social groups been assembled themselves and defended their causes. On 04/18/2016, the Brazilian magazine Veja published an article entitled “bela, recatada e do lar” (beautiful, demure and housewife), whose repercussion mobilized several social groups and generated a virtual protest in nationwide scale. The goal of this research was to analyze behavior of users in the social network Twitter to identify how people reacted to the article “bela, recatada e do lar”. To achieve this, a network of shared messages (retweets) was built, where the centrality metrics Degree, Betweenness and Pagerank were calculated to identify which users most influenced the social movement. Also, a data mining technique known as sentiment analysis was used with the aid of the ETL (Extract, Transform & Load) methodology and the Naive Bayes probabilistic algorithm to study users behavior and opinion. Furthermore, an analysis of highlighted events was performed from most frequently tweeted hashtags. Results showed that (i) users that had more influence in the social movement could be split into two main classes: one represented by users with high Pagerank values, or in other words, users that published relevant content and were shared extensively by others, and another class represented by users with high Betweenness values, meaning that they acted in an influential manner only inside specific communities; (ii) in its majority, users expressed opinions against the conservative standard for women defended by the magazine article; and (iii) events that occurred in parallel to the social movement “bela, recatada e do lar” apparently influenced the content and amount of published messages in the OSN.