A study conducted by a team of researchers from Instituto Superior Técnico addressing specificities of online hate speech against Afro-descendant, Roma and LGBTQ+ communities was published in June, in the Journal of Language Aggression and Conflict. The study aims to analyse the prevalence and linguistic strategies underlying online hate speech against these communities. The research was developed within the framework of the HATE COVID-19.PT project and aims to develop automatic recognition systems to detect this type of speech.
The team of researchers is led by Técnico professor Paula Carvalho, and includes researchers from the Instituto de Engenharia de Sistemas e Computadores (INESC-ID) and the Interactive Technologies Institute (ITI-LarSyS).
The researchers created a corpus composed of more than 20,000 YouTube comments posted by 8,485 online users on 39 YouTube videos targeting those communities. The comments were meticulously analysed using linguistic techniques, which allowed the identification of explicit and covert hate speech, counter-speech, and offensive speech patterns.
The creation of the first annotated corpus for European Portuguese will be a valuable resource for studying and detecting online hate speech, especially with regard to targeted communities, on social media platforms.
