| dc.description.abstract | In this chapter, which complements the chapter on “AI & Law: Logical Models,” we consider AI & law research on case-based reasoning and machine learning. The two approaches share a focus on legal data concerning individual decisions, rather than general rules and concepts. Both aim at extracting useful information from past cases. In some, but not all application, the extracted information concerns ways to address new cases. However, the two approaches, as they have been developed within AI & law, exhibit a significant difference. Case-based reasoning has relied on the human encoding of information on individual cases, to be processed according to reasoning moves meant to extend the outcome of past cases to new ones, or rather to distinguish the new case from the past ones. The machine-learning approach on the contrary, has focused on the automated construction of models, often having a non-symbolic nature. Connection between the two approaches have emerged, in particular, through the use of machine learning models to extract information to be employed in case-based reasoning. | en |