Research Projects

The COVID-19 pandemic has led to a surge in the use of social media, creating a large amount of user-generated textual data. Artificial intelligence approaches to textual analysis use deep neural networks, which require text to be transformed into a numeric form through text embedding. This allows for cross-lingual knowledge transfer, improving low resource language processing. The project aims to construct cross-lingual embeddings for nonstandard languages, such as Hindi and Slovene, to enable efficient transfer of analytical models for sentiment and emotion. The developed tools will automatically detect hate speech generated by bots and adversarial users, making classifications public and reducing harmful content. The project will involve establishing an advanced technological basis for understanding non-standard internet languages, developing location-sensitive sentiment and emotion analysis of tweets and online comments, and analyzing hate speech specific to the COVID-19 epidemic. The project will also provide trending topics of concern and develop an anti-hate speech bot to warn human hate speech spreaders of the legal implications of their actions.