Development of chatbot to support student learning strategies in design education

Abstract

We have developed a chatbot that advises design students on their learning strategies through automatic interactions. We implemented a prototype of the chatbot in a sketchbook application that students utilized to receive advice during their design studies and evaluated whether it could help them to facilitate their learning strategies. The results demonstrated that, while the chatbot was generally able to create a natural dialogue flow with the students and helped to increase some learning strategies, there was still significant room for improvement. We analyzed the data on learning strategies and self-efficacy obtained from the experiments in detail and came up with a new hypothesis. That is, the experimental result indicated that chatbots can increase strategies for understanding learning progress, reflecting on learning, and devising learning methods by interacting with students in a way that gives them confidence, and we will empirically verify it in our future work.