Classifier Bias Detector

A bias detector that applies a machine-learning classifier trained on labeled biased and neutral text to recognise implicit discriminatory language patterns.

Responsibility. Classifies text as biased or neutral using a trained bias classifier.

Also known as: ML bias classifier, Content moderation classifier

Variant of Output Bias Detector abstract

When to choose. Choose for robust, scalable detection on user-facing content, including implicit bias without explicit markers; requires labeled training examples and periodic retraining.