Abstract
Remote learning technology has advanced both in online videos and Zoom classes. In providing high quality, interactive learning experiences, the rate limiting step is often capacity of the teacher operating the Zoom class, or answering questions arising from videos. This project seeks to address this limitation by supporting the human intelligence of teachers with Artificial Intelligence (AI) conversational agents who can give each and every student an interactive experience similar to that of a personal teacher.
In our project, we will use adapt existing cloud-AI technology to create chatbots which students interact with just like chatting on a platform like Facebook. Medical students reviewing video learning materials replacing anatomy dissection classes will be supported by the chatbot, which both asks them questions and answers their questions. This will address an otherwise impossible task: the need for teachers to give instantaneous feedback to all students, whilst simultaneously gathering data on their performance.
Critically, as AI agents they will improve as we gather more data from student-AI interactions. Unlike new software which becomes outdated over time, our AI improves as this database increases. This enforced period of remote learning is the perfect environment to kick-start the initial development of sophisticated AI agents and launch our students into the era of AI-supported learning.
Brief write-up
Project objectives
• To provide interactive support to students undertaking the MBChB Year Two “Human Structure I” (MEDU2300) online video learning, via an autonomous AI chatbot.
• To evaluate the effectiveness of the AI chatbot in supporting learning.
• To build a starting point to develop further educational AI support for students.
Activities, process and outcomes
• Designed an architecture for an Artificial Intelligence Chatbot for teaching human anatomy for medical students and deployed via our own web-app in order to deliver chatbot teaching.
• Both “Human Structure I” (MEDU2300) and “Human Structure II” (MEDU3300) students were offered the use of this chatbot in support of their learning in pilot settings.
Deliverables and evaluation
1. Successful development and deployment of sophisticated AI chatbot for teaching and learning.
2. Supported student learning of anatomy for medical Year Two and Three students with positive qualitative rating and qualitative comments.
3. Influenced local practice, delivering one CLEAR (Centre for Learning Enhancement And Research) webinar on AI chatbots (29/3/2021) and further scholarly dissemination (below).
Dissemination, diffusion, impact and sharing of good practices
Disseminated findings in two conferences (one local and one international), receiving two awards and one journal publication accepted.
Impact on teaching and learning
1. Students were provided with an additional tool to further their learning in a pilot laboratory setting – to stimulate educational conversation and both ask and answer questions.
2. Student users rated our tool very positively, including ratings of 5.12/6 (Enjoyed the use of chatbot), 4.96/6 (Would recommend to others to use).
3. Teachers have responded very positively this initiative and will further develop its use.
