Recent Research from the NSF AI Institute for Adult Learning and Online Education (AI-ALOE)

• In Event: Next-Generation Educational AI Research From Three NSF-Funded National AI Institutes

Abstract

Objectives
AI-ALOE develops learning and teaching assistants for online education for adult learners, for example, intelligent textbooks, smart videos, and agent-based simulations for enhancing cognitive engagement, teaching presence, and social interactions. We conduct foundational research on personalization of learning and human-AI interaction in the learning and teaching assistants we develop.

Theoretical framework
AI-ALOE research on online education for adult learners is guided by a range of theoretical perspectives including Community of Inquiry, adaptive and personalized learning at scale, machine teaching and trainable agents, theory of mind for generating self-explanations, and mutual theory of mind for human-AI communication.

Methods and Data Sources
We adopt the methodology of learning engineering that entails theory-based design, human-centered design, participatory design for human well-being, intervention in real-world educational settings, extensive formative assessment, and iterative and sustained data-driven revision of theories and designs. We have introduced six AI assistants for enhancing cognitive, teacher, and social presence in about fifty classes, mostly at the Technical College System of Georgia which has a very diverse population of adult learners: male and female, majority and minority, urban and rural, civilian and military, and neurotypical and neurodiverse.

Results
Results from our research include theories, techniques, and tools for enhancing cognitive engagement, teaching presence, and social interactions in online education for adult learners; personalization in skill learning and concept learning; machine teaching for constructing teachable agents; meta-reasoning for self-explanation; and integration of cognitive and generative AI for constructing learning and teaching assistants.

Authors. (2024) Social AI Agents Too Need to Explain Themselves. In Procs. 20th International Conference on Intelligent Tutoring Systems..
Authors. (2024). SAMI: An AI actor for fostering social interactions in online classrooms. In Procs. 20th International Conference on Intelligent Tutoring Systems.
Authors. (2024). Investigating the influence of AIaugmented summarization on concept Learning, summarization skills, argumentative essays, and course outcomes in online adult Education. In Procs. 18th International Conference of the Learning Sciences.
Authors. VAL: Interactive Task learning with GPT dialog parsing. In Procs. 2024 CHI Conference on Human Factors in Computing Systems.
Authors. (2024). Formative feedback on student-authored summaries in intelligent textbooks using large language models. International Journal of Artificial Intelligence in Education.
Authors. (2024) Jill Watson: A virtual teaching assistant powered by ChatGPT. In Procs. 24th International Conference on AI in Education.

Author

• Ashok Goel, Georgia Institute of Technology