Font Size: 

It is gradually becoming decisive to address the implications on Italian and European society of our behavior towards immigration. Rather than source of cultural enhancement, this phenomenon is often related to prejudice and more or less direct feeling of hostility to foreign people, frequently source of hate speech in communication media. The paper describes an ongoing project for the development of artificial intelligence tools and resources for detecting and analyzing hate speech and prejudice against immigrants in social media. It is the result of a multi-disciplinary team that includes skills from computational linguistics and sentiment analysis, to social network analysis and visualization. The approach follows a multi-lingual perspective, which encompasses different knowledge sources to be meaningfully aggregated. 
We propose to integrate innovative techniques for collecting big social data and applying to them data mining and aspect-based sentiment analysis, 
i.e. highly developed natural language processing to deal with the fine-grained aspects involved in hate speech and hostility to immigrants. A geography of hate and prejudice about immigrants and a study of immigrants mobility across our territory will be derived from geo-social data. The aim is twofold: transforming analyzed data in actionable knowledge for local entities; democratizing data.