Use of Natural Language Processing to Classify Open-Ended Responses on Course Evaluation Questionnaires at Rīga Stradiņš University

Description

Student feedback is used to continuously improve the study process at Riga Stradiņš University (RSU). Student responses on Course Evaluations Questionnaires is analysed by various stakeholders (lecturers, study programme directors, university administration etc.) and improvements to multiple aspects of the study process (lectures, classes, assessment, learning materials etc.) are made on several levels (course, study programme or institutional level). One of the challenges in obtaining actionable information from the student feedback on Course Evaluation Questionnaires is the amount of work required to systematically analyse student responses to the open-ended questions. This research developed and evaluated a Natural Language Processing solution to support the analysis of such text based feedback.. To obtain student feedback on the study process, RSU uses online study course evaluation questionnaires. Students evaluate multiple aspects of course quality using closed and open-ended questions. In the Spring semester of 2022, students submitted 13 thousand completed questionnaires with more than 9 thousand open-ended responses. A semi-automated classification system was developed by the RSU Centre for Educational Growth to label student responses on open-ended questions with meaningful categories, for example, “Assessments”, “Independent Work”, “Lectures” based on keywords used in the student response text. Until year 2023 this classification was done manually. To improve this process RSU Department of Information Technology developed a machine learning algorithm to automate and extend the classification of student responses.. Use of Natural Language Processing methods allows a more precise and faster division of student responses into pre-defined categories. During the training process a machine learning algorithm also identified new keywords for labelling student responses, extending the previously used classification schema.. Natural Language Processing and machine learning algorithms can be used to enhance study quality by supporting acquisition of actionable insights from open-ended student feedback on the study process and outcomes.

Period	30 Mar 2023
Event title	RSU research week 2023: University Teaching and Learning
Event type	Conference
Organiser	Rīga Stradiņš University
Location	Riga, LatviaShow on map
Degree of Recognition	International