Curiosity Detection in Student Text Data : An Empirical Investigation for a Computer Science Course

Abstract

Instilling curiosity in students improves their learning process. How can we understand the degree of curiosity among students in a specific course? The Question Formulation Technique (QFT) lends itself toward understanding a topic-specific curious mind. In a senior-level course, we collect data using the QFT model, which is then analyzed using Natural Language Processing (NLP). Thought provoking statements are given to students to analyze, who then respond to them with answers in the form of questions. Python scripts are programmed to analyze the student responses and the WEKA data mining tool is used for feature (words) extraction and classification. We conclude that the features extracted provide excellent insight into “Propensity for Exploration” (PE) as a measure of curiosity in student text data.