Fighting offensive language on social media with unsupervised text style transfer

Online social media has become one of the most important ways to communicate and exchange ideas. Unfortunately, the discourse is often crippled by abusive language that can have damaging effects on social media users. Online social media networks normally deal with the offensive language problem by simply filtering out a post when it is flagged as offensive. In the paper "Fighting Offensive Language on Social Media with Unsupervised Text Style Transfer," which was presented in the 56th Annual Meeting of the Association for Computational Linguistics (ACL 2018), we introduce a completely new approach to tackle this problem. Our approach uses unsupervised text style transfer to translate offensive sentences into corresponding non-offensive forms.