Introduction to Reinforcement Learning
Introduction to Reinforcement Learning
| Hi-index | 0.00 |
We have developed a tool for generating hints within computer-aided instructional tools based on a corpus of student work. This tool allows us to select source problem solutions that match the current user solution and generate hints based on next problem steps that are most likely to lead to a successful solution. However, within such a tool it is possible to generate hints that did not turn out to be useful in the source problem solution. Therefore, we have proposed a metric to measure and integrate a “utility” function to choose source material for hint generation. In this paper we present our metric and an experiment to investigate its use on real data from a logic proof tutorial.