LabHC/moji

The Moji dataset (Blodgett et al., 2016) (slanglab.cs.umass.edu) contains tweets used for sentiment analysis (either positive or negative sentiment), with additional information on the type of English used in the tweets which is a sensitive attribute considered in fairness-aware approaches (African-American English (AAE) or Standard-American English (SAE)). The type of language is determined thanks to a supervised model. Only the data