An Exploration of Data-Driven Hint Generation in an Open-Ended Programming Problem

Data-driven systems can provide automated feedback in the form of hints to students working in problem solving environments. Programming problems present a unique challenge to these systems, in part because of the many ways in which a single program can be written. This paper reviews current strategies for generating data-driven hints for programming problems and examines their applicability to larger, more open-ended problems, with multiple, loosely ordered goals. We use this analysis to suggest directions for future work to generate hints for these problems.