Toward supporting hypothesis formation and testing in an interpretive domain

The research field of AI & Education has long been interested in cognitive processes in which students formulate and test hypotheses by considering them in light of specific cases. However, few if any of the systems that have been built target domains which are ill-structured and in which determining whether a hypothesized rule and proposed outcome are consistent with past decisions is a matter of interpretation, rather than deductive inference. The goals of our project are to (1) develop an AI model of hypothesis formation and testing in an interpretive domain, US Supreme Court oral arguments and (2) to use it in an intelligent tutoring system to guide law students in learning that process. As a first step toward these goals we will conduct an experiment to evaluate whether self-explanation prompts facilitate learning by studying argument transcripts.