Abstract:
Large-scale lectures with more than 50 students per teacher and MOOCs with thousands of learners are common in higher education. It is impossible for one single teacher, even with the help of a tutor, to provide individualized support in such cases. Although one-on-one individualized tutoring can help provide personalized study tips to enhance students’ learning performance and experience if timely feedback is received, it is difficult to be employed in large class sizes or intensive courses.
In this project, we proposed using an AI chatbot as an intelligent tutor as a potential solution. This approach compensates the insufficient individual support in lectures and provides an adaptive learning environment to meet the needs of a large number of diverse learners. Personalized reading recommendation will be provided based on students’ learning progress and performance in the formative assessments. Students can learn about their learning and achievement and obtain immediate feedback without time and location constraints during the conversation in the chatbot.
To study the effect of using AI chatbots as intelligent private tutors on the learning and teaching experiences of both students and teachers in higher education, experiments were conducted in four undergraduate courses in EdUHK and PolyU.
Code:
T0225
Principal Project Supervisors:
Keywords provided by authors:
Start Date:
25 May 2020
End Date:
24 Nov 2021
Status:
Completed
Result:
The AI chatbot was launched and integrated into an online evidence-based assessment system – GMoodle (https://gmoodle.eduhk.hk) – in September 2020, adding to the base functionalities of the system. This TDG project includes three phases (i.e. semester 1 and 2 of 2020/21 and semester 1 of 2021/22). In each phase, two courses were selected to use this online platform to promote personalized learning with AI chatbot, and there were 295 students from undergraduate and postgraduate programmes.
Most of the students agree that GMoodle with AI chatbot helps to provide personalized support in their learning. They can communicate more efficiently, had better preparation for class and had put a great deal of effort into advancing their learning in the courses. Overall, the self-evaluation scores are all increased after using the GMoodle platform as shown below. In the focus group interview, students reflected that they would like to express their opinion more in class than before, and they got timely feedback from the teachers and the AI chatbot which helps them to save time to search for the required information. Furthermore, the personalized feedback can help them understand their study progress and level. In the teacher perspective, this intelligent online learning platform can help catering to learning diversity and provide customized feedback to the students.
Impact:
The impact of the AI chatbot can be referenced to the pre-test/post-test survey. There were high percentage of students who agree or strongly agree with the corresponding aspects are shown in the figure below. This preliminary result showed that GMoodle with AI chatbot can help the students to improve their group communication, peer understanding and identification of common learning goals. Besides, they can enhance the problem-solving skills and self-evaluation skills.
There is also a growing need in self-regulated learning. This online platform can help the teachers to understand students’ needs and progress to provide personalized recommendations to the students. Other than using the AI chatbot to answer some common questions, students were encouraged to explore the course by referencing to feedback generated by the AI chatbot. The students can seek for help 24-hour without worrying about the frequency of asking questions. Most of the students enjoyed the learning process with the extra support from the AI chatbot.
Deliverables:
Books/ Book Chapters/ Journal Articles/ Conference Papers
Kong, S. C., Lam, S. M. S., Lam, W. M. W., LAU, K. M. K., YEE, T. S. L. & Yung, S. N. R (2021, November 8-9). Four case studies on applying alternative assessment in real situations with technology tools to facilitate teaching and learning [Conference presentation]. 14th annual International Conference of Education, Research and Innovation, Spain. (Number of participants: 100) (https://repository.eduhk.hk/en/publications/four-case-studies-on-applying-alternative-assessments-in-higher-e)
Lam, W. M. W., Liu, Y. W. D. & Yung, W. H. K. (2020, November 9-11). Providing Personalized Learning Support with AI Chatbot. International Conference on Education and Artificial Intelligence 2020. The Education University of Hong Kong. (Number of participants: 80) (https://repository.eduhk.hk/en/publications/providing-personalized-learning-support-with-ai-chatbot)
Seminars/Presentations/Sharing Sessions
Lam, W. M. W. (2021, December 7). Providing Personalized Feedback and Promoting Active Learning Experience with AI Chatbot on Moodle-based Learning Platform. The Hong Kong Polytechnic University. (Number of participants: 20) (https://repository.eduhk.hk/en/publications/providing-personalized-feedback-and-promoting-active-learning-exp)
Lam, W. M. W. (2022, March 24). Alternative Assessment Strategies to Promote Collaborative and Self-regulated Learning in Group Activities. Teaching and Learning Community of Practice Sharing Series - Assessment Strategies for Online Teaching. The Chinese University of Hong Kong. (Number of participants: 30) (https://repository.eduhk.hk/en/publications/alternative-assessment-strategies-to-promote-collaborative-and-se)
Teaching and Learning Resources/ Materials (including online resources)
GMoodle with AI Chatbot. (https://gmoodle.eduhk.hk/ ) (https://repository.eduhk.hk/en/publications/gmoodle-with-ai-chatbot)
Financial Year:
2019-20
Type:
TDG