Memory Networks

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.