Real-Time Competency Mapping of Student Reflections Using Rule-Based AI and Visual Analytics

Abstract

The analysis of qualitative data at scale remains a major challenge in educational research, particularly when student reflections are unstructured and open-ended. This study proposes a symbolic artificial intelligence (AI) approach using rule-based natural language processing (NLP) to automate the classification of qualitative text data. Specifically, 400 project reflections from graduate students in supply chain and operations courses were processed and mapped to eleven predefined competency categories. By converting free-text reflections into structured binary classifications, this system enables scalable, interpretable, and real-time analysis of qualitative learning evidence. This method eliminates the need for manual coding while preserving transparency, making it especially useful in educational settings where explainability is critical. Visual analytics—including frequency distributions, word clouds, bubble charts, and correlation heatmaps—were generated to support insight extraction. Results demonstrate that rule-based NLP can effectively bridge the gap between qualitative input and structured analysis, offering a practical solution for research, assessment, and program evaluation where traditional machine learning may be too opaque or resource-intensive.