Computer-based Classification of Student's Report

Abstract. Learning Analytics focuses on measuring and analyzing learners’ data, such as formative assessment of collaborative writing and individual students’ performance. This work applied machine learning approaches and natural language processing to assess university students’ reports in the knowledge building domain. Students (n = 32) wrote essays about knowledge building topics, and the professor used Dublin descriptors as assessment criteria to evaluate the students’ reports. This paper presents the results of a study on validating whether students’ reports are aligned with Dublin descriptors for qualifications awarded. We have used two classification models: Support Vector Machine (SVM) and Random Forest Classifier (RFC), to predict manual annotations from experts in students’ reports. Random Forest Classifier reached 73% accuracy. We concluded that machine learning algorithms and natural language processing (NLP) together are useful for automating the classification of the students’ reports using manual annotations.