Analysis: “Speech containing offence, prejudice and hate against Roma on the social media platform Facebook”
Materials were collected between September 1 and November 30, 2023. Out of the 1,252 texts mentioning Roma in some form, at least 80% have offensive content towards Roma, but if we do not count content posted in Roma-only groups on social media or by Roma organisations, the share of negative posts would actually exceed 90%. The Facebook search was done automatically using keywords pre-selected by experts. The posts were analysed experimentally using artificial intelligence.
The methodology and the algorithms used for automatic analysis were developed by a consortium of organisations from the MRG, ILC and the Amalipe Centre with funding from the EU. Textgain, a language technology company set up by the University of Antwerp, was selected as the company to collect and analyse the data using artificial intelligence. The partner for the collection and processing of the data in Bulgarian is Center Amalipe.
This report is based on semiautomated analysis of the collected corpus of posted material.