Preliminary Findings of using Chat-bots as a Course FAQ Tool

Abstract

Students learn by asking questions and as instructors we encourage students to ask questions. However, not all questions are of equal importance. Questions related to the mechanics of a course are unavoidable but when the same question is asked and answered multiple times by an instructor it can become a burden to the instructor to personally answer these questions multiple times. This paper presents the preliminary findings of an admin-updatable Dialogflow-based chat-bot deployed on an instant messaging platform (Telegram) to handle all course queries. The chat-bot uses the in-built natural language processing module by Dialogflow to understand a student's query and match it to the provided predefined answers. The method outline in this paper describes the process to create a chat-bot from a limited set of course related questions, use the chat-bot as a data collection tool to create a question bank and use the new questions to improve the functionality of the chat-bot to support the course. Our observations and findings from the first iteration of deployment shows that chat-bots are a viable tool to disseminate information to students and it is possible to build such an agent with a limited set of question.