Mitigating Racial Biases in Toxic Language Detection with an Equity-Based Ensemble Framework.
Matan HalevyCamille HarrisAmy BruckmanDiyi YangAyanna M. HowardPublished in: EAAMO (2021)
Keyphrases
- main contribution
- united states
- high level
- conceptual framework
- language learning
- declarative language
- context dependent
- programming language
- detection algorithm
- detection method
- strongly typed
- data sets
- specification language
- detection accuracy
- ensemble methods
- false positives
- object detection
- software engineering
- probabilistic model
- natural language
- neural network