Not all swear words are used equal: Attention over word n-grams for abusive language identification

Abstract

The increasing propagation of abusive language in social media is a major concern for supplier companies and governments because of its negative social impact. A large number of methods have been developed for its automatic identification, ranging from dictionary-based methods to sophisticated deep learning approaches. A common problem in all these methods is to distinguish the offensive use of swear words from their everyday and humorous usage. To tackle this particular issue we propose an attention-based neural network architecture that captures the word n-grams importance according to their context. The obtained results in four standard collections from Twitter and Facebook are encouraging, they outperform the $$F:1$$ scores from state-of-the-art methods and allow identifying a set of inherently offensive swear words, and others in which its interpretation depends on its context.