Educational data mining: A case study of teacher's classroom questions

Abstract

This paper presents a new application of data mining techniques, particularly text mining, to analyze educational questions asked by teachers in classrooms. More specifically, it reports on the performance of four machine learning techniques and four feature selection approaches on the classification of teacher's questions into different cognitive levels identified in Bloom's taxonomy. In doing so, a dataset of questions has been collected and classified manually into Bloom's cognitive levels. Preprocessing steps have been applied to convert questions into a suitable representation. Using the dataset, the performance of machine learning techniques under feature selection approaches has been evaluated. The results show that Rocchio Algorithm performs the best regardless of the used feature selection approach. Moreover the best RA performance can be obtained when Information Gain is used for feature selection.