Knowledge Graph Representation of Felder-Silverman Learning Style Model for Computing Education

This research focuses on adapting the Felder-Silverman Learning Style Model (FSLSM) questionnaire according to a specific computing course. In this approach, we transform FSLSM questions based on the course’s algorithmic structure. Using these transformed questions, we create Learning Style Knowledge Graphs (LSKGs), where each learner has their own LSKG based on their learning style. Once these LSKGs are generated, we use them to form clusters of learners who share similar features. These clustered learners can enhance their learning experience through their shared learning profiles, as they exhibit similar characteristics. To increase learners’ engagement with the questionnaire, we have made the questions interactive and aligned with the algorithmic course. Our analysis indicates that around 26 clusters are formed, which, according to Silhouette and Calinski-Harabasz scores, provide the best clustering results. We observed that when the threshold is set above 0.5, approximately 76% of learners have matching profiles with each other. This highlights the need to consider individual learning styles when designing educational strategies. The LSKG approach can be used to pinpoint students’ strengths and weaknesses, providing a foundation for AI-based adaptive learning systems that tailor educational content to the needs of each student. Moreover, LSKGs enable effective resource allocation by aligning teaching methods with learning styles.