Explainability

Detecting and explaining unfairness in consumer contracts through memory networks

We posit that in this domain useful explanations of classifier outcomes can be provided by resorting to legal rationales.