AI-Infused Semantic Model to Enrich and Expand Programming Question Generation

Creating practice questions for programming learning is not easy. It requires the instructor to diligently organize heterogeneous resources, i.e., conceptual concepts and procedural rules. Today’s question generation (PQG) still largely replying on demanding creation task performed by instructors without advanced technological support. In this work, we propose a semantic PQG model that aims help generate new expand assessment items. The designed transform knowledge from textbooks into network Local Knowledge Graph (LKG) Abstract Syntax Tree (AST). For any given question, queries established find related code examples generates set of associated LKG/AST structures. We conduct analysis compare instructor-made 9 undergraduate introductory courses textbook questions. results show had much simpler complexity than ones. disparity topic distribution intrigued us further research breadth depth quality also investigate in relations student performances. Finally, report an user study proposed AI-infused examining machine-generated quality.