Author(s):
With the popularization of artificial intelligence tools in day-to-day processes, it is important to examine and test how the technologies can be used to improve the education environment. Virtual assistants, aka chatbots, can be designed as a versatile classroom aid, easing the burden on the instruction staff and increasing information accessibility to students. In particular, building curated chatbots that fit classroom needs with verified information while leveraging machine learning and natural language processing capabilities with minimal fuss makes for an attractive use case. Initially developed for large lecture-style, introductory college-level chemistry classrooms, the curated chatbots have been iterated for various uses in a number of different chemistry classroom spaces (25-400 students). They have been designed to provide information about writing chemistry in context, prepare students for laboratory experiments, and respond quickly to utilitarian information about the course, such as classroom policies and deadlines. Routinely, the curated chatbot reached 30-80% of students over the last two years. How the students used the curated chatbot depended on the course and the types of queries posited to the chatbot. An overview of the curated chatbot’s development process, use cases, analysis, and student use patterns are discussed.
Coauthors
Annabelle Lolinco, Iowa State University, Ames, IA
