Landmarks in Case-Based Reasoning: From Theory to Data

Widespread application of uninterpretable machine learning systems for sensitive purposes has spurred research into elucidating the decision making process these systems. These efforts have their background in many different disciplines, one which is field AI & law. In particular, recent works observed that training data can be interpreted as legal cases. Under this interpretation formalism developed to study case law, called theory precedential constraint, used analyze way draw on – or should them make decisions. predominantly stay theoretical level, hence present work evaluated a real world dataset. Through analysis we identify significant new concept call landmark cases, and use it characterize types datasets are more less suitable described by theory.