Emotional text mining of social networks: the French pre-electoral sentiment on migration

Migration has actually gained considerable relevance both in national and European political agendas and in general public debate. The migratory phenomenon, as well as its humanitarian and health relevance, is presented nowadays as a challenge for national and supranational governments, which needs coordinated responses to ensure citizens' security (Coppola & Macioti, 2017). The issue of security is also related to the terrorist attacks of religious matrix perpetrated on western countries since the beginning of the new millennium that have shaken public opinion, highlighting the issue of freedom of movement and residence for people within the European Union. During the presidential elections in France, the populist rhetoric had largely exploited the perception of the freedom of movement as a risk factor, taking advantage of this issue in the political propaganda. The Front National, a far-right party led by Marine Le Pen, who was considered one of the favourite presidential candidates, quadrupled its consensus getting 25% of the vote in the European elections of 2014, focusing its political campaign on topics such as border closure and exiting from the EU. In recent years, social media analysis has become a fast and cheap device, compared to the traditional survey, to explore the political and electoral opinions and sentiments of citizens. Moreover, social network analysis was used for several purposes, such as demonstrations and revolt organization, the engagement of individuals in mobilization, and the construction of social movements and political parties (e.g. the Cinque Stelle Italian political party). For this reason, social media and social network sites, like Facebook and Twitter, have started to play a growing role in real-world politics (Ceron, Curini, Iacus & Porro, 2014). The wide diffusion of the internet increases the opportunity for millions of people to surf the web, create account profiles and search or share information daily. The constant rise in the number of users of social media platforms, such as Twitter, make a large amount of data available that represents one of the primary sources to explore people’s opinions, sentiments, and emotions (Ceron, Curini & Iacus, 2013; Pelagalli, Greco & De Santis, 2017). Therefore, texts can be analysed in order to explain and anticipate the dynamics of different events, such as stock market activity, elections, etc. (e.g., Schoen et al., 2013; Ceron, Curini, Iacus & Porro, 2014), potentially producing useful results applicable in different contexts. There are a variety of procedures used to extract such information from different types of textual data focusing on several procedures as shown by the literature (Reinert, 1983; Halfon et al., 2016; Hopkins and King, 2010; Ceron, Curini & Iacus, 2016; Pelagalli, Greco & De Santis, 2017). In this paper, we analyse the sentiment about migration in social media during the French presidential campaign of 2017. We perform an emotional text mining (Cordella, Greco & Raso, 2014; Greco, 2016; Pelagalli, Greco & De Santis, 2017) in order to explore the emotional content of the Twitter messages concerning migration written in French in the last two weeks before the first round of the presidential election on April 23th, 2017. The aim is to analyse the opinions, feelings and shared comments, classifying the contents and measuring the sentiments. This procedure allows for the detection of the emotional representation of migration emerging from tweets in the pre-election period. The paper is composed of 5 sections, as follows: in section 2, we describe the methodology of the emotional text mining; in section 3, we describe the collection and analysis of data; in section 4, we illustrate the main results; in section 5, we discuss the conclusions.