Rejection and hate speech in Twitter: Content analysis of tweets about migrants and refugees in Spanish

Abstract

We use Twitter to study the verbal rejection towards migrants and refugees as a potential hate speech predictor with two content analysis of tweets in Spanish collected with Twitter's API: the first analysis, manual, with 1,469 tweets; the second, automatic, uses big data techniques to study 337,116 new tweets. In the first one rejection was predominant over acceptance and neutrality. Rejection was smaller in the second one, showing how fluctuant these expressions are depending the media context. In both cases rejection toward migrants was significantly bigger than over refugees, as it had already been observed in international contexts. This work also created a training corpus about immigrant rejection, valid for future studies, and observed the negative aspects most frequently associated to rejection of immigrants, as well as the relationship existing between this and the fact of tweets being information or opinion.