Uncovering Patterns in Student Exam Feedback: An NLP Approach (Data)

This restricted-access record contains the dataset used in the B.Sc. thesis “Uncovering Patterns in Student Exam Feedback: An NLP Approach.” The data consist of student exam complaints and corresponding staff responses, complaint statuses, exercise-level scores before and after complaint handling, exam-result exports, mapping information, and exam-specific grading schemas. These data supported preprocessing, exploratory BERTopic topic modeling, hybrid zero-shot and rule-based classification, text-feature extraction, validation, descriptive association analysis, and exam-context analysis. Public access is restricted because the deposited records contain sensitive educational and administrative information. Access may be granted only following authorization by the responsible institution and in accordance with applicable data-protection requirements.