Building a generic feedback system for rule-based problems

Abstract

We present a generic framework that provides hints on how to achieve a goal to users of software supporting rule-based problem solving from different domains. Our approach consists of two parts. First, we present a DSL that relates and unifies different rule-based problems. Second, we use generic search algorithms to solve various kinds of problems. This solution can then be used to calculate a hint for the user. We present three rule-based problem frameworks to illustrate our approach: the Ideas framework, PuzzleScript and iTasks. By taking real world examples from these three example frameworks and instantiating feedback systems for them, we validate our approach.