Interactive correction and recommendation for computer language learning and training

Abstract

Active learning and training is a particularly effective form of education. In various domains, skills are equally important to knowledge. We present an automated learning and skills training system for a database programming environment that promotes procedural knowledge acquisition and skills training. The system provides meaningful knowledge-level feedback such as correction of student solutions and personalized guidance through recommendations. Specifically, we address automated synchronous feedback and recommendations based on personalized performance assessment. At the core of the tutoring system is a pattern-based error classification and correction component that analyzes student input in order to provide immediate feedback and in order to diagnose student weaknesses and suggest further study material. A syntax-driven approach based on grammars and syntax trees provides the solution for a semantic analysis technique. Syntax tree abstractions and comparison techniques based on equivalence rules and pattern matching are specific approaches. © 2006 IEEE.