Enriching programming content semantics: An evaluation of visual analytics approach

Abstract

In this work, we present an intelligent classroom orchestration technology to capture semantic learning analytics from paper-based programming exams. We design and study an innovative visual analytics system, EduAnalysis, to support programming content semantics extraction and analysis. EduAnalysis indexes each programming exam question to a set of concepts based on the ontology. It utilizes automatic indexing algorithm and interactive visualization interfaces to establish the concepts and questions associations. We collect the indexing ground truths of the targeted set from teachers and experts from the crowd. We found that the system significantly extracted more and diverse concepts from exams and achieved high coherence within exam. We also discovered that indexing effectiveness was especially prevalent for complex content. Overall, the semantic enriching approach for programming problems reveals systematic learning analytics from the paper exams.