"Our Mission is to Build on Theories of Learning and Instruction to Create Innovative Learning Environments that Maximize Learner Capacity to Achieve Learning Goals"
AI2 lab members attended the AI ALOE's Year 2 Executive Advisory Board Meeting
May 19, 2023
Director Dr. Min Kyu Kim and the graduate associates from the AI2 lab, Jinho Kim, Lia Haddadian, and Yoojin Bae, attended the AI ALOE's Year 2 Executive Advisory Board Meeting on May 19, 2023. Approximately 40 members, including 7 EAB members, participated in the meeting either in person or online.
During the meeting, Dr. Kim presented his recent AI in education research titled "Personalized Learning Strategies for Concept Learning."
Additionally, Jinho and Yoojin participated in the poster session, where they presented their recent research work:
For more information about the EAB meeting, please visit the following link: https://aialoe.org/eab2023/
NSF AI ALOE Institute Research Videos Now Available
May 4, 2023
We have created two video presentations for public access. The videos introduce our AI research and development, centering on two topics: Concept Learning and Personalization. We define concept learning as the cognitive process of building a solid understanding of STEM content necessary for developing critical thinking and problem-solving skills. Personalization is one of the goals pursued in AI in education that enables learners to be better engaged in and regulate their learning and performance. Additionally, personalized support for instructors can empower them to support learners in more adaptive and appropriate ways. Our three graduate associates, Jinho, Lia, and Yoojin, have served as narrators. Enjoy!
Attending and Presenting at the 2023 AERA Annual Meeting
April 13, 2023
Members of our AI2 Lab attended and presented at the 2023 American Educational Research Association (AERA) Annual Meeting, which took place from April 13-16 at Chicago, IL.
Dr. Min Kyu Kim and our graduate associate, Jinho Kim, presented the paper titled "Using Machine Learning for Cognitive Presence Detection in Asynchronous Online Learning" at the Instructional Technology SIG Roundtables on the 13th.
Using Machine Learning for Cognitive Presence Detection in Asynchronous Online Learning
This study aims to explore and develop a machine learning model to automatically detect individual students’ levels of cognitive presence they exert while participating in online discussions. To this end, we used the Practical Inquiry Model to analyze 1,360 discussion posts gathered from the multiple sections of a graduate-level online course. We have used the training data to develop a machine learning model. Especially, we tested a deep learning algorithm, Bidirectional Encoder Representations from Transformers (BERT) model, to find the most suitable hyperparameters. Our final model showed a 94% accuracy in detecting cognitive presence levels reflected on given discussion posts. Findings demonstrate the potential of AI to detect students’ cognitive presence and provide real-time feedback.
Yoojin presented her recent study at LS-GSA Spring Student Conference
April 7, 2023
Our graduate associate, Yoojin Bae, presented her recent study entitled "Clustering Cognitive Engagement Changes in Longitudinally Traced Discussion Data from a Graduate-Level Online Course" at the 2023 Learning Science (LS)-Graduate Student Association (GSA) Spring Conference. All of our lab members attended and celebrated her first but wildly successful presentation. Yoojin is in her second semester.
Clustering Cognitive Engagement Changes in Longitudinally Traced Discussion Data from a Graduate-Level Online Course
Given the pervasive use of online learning, especially asynchronous discussion forums, ensuring students’ ongoing engagement is critical. Examining students’ cognitive presence throughout a series of discussions can disclose different engagement patterns. However, little research has leveraged longitudinal discussion data. The current study aims to identify learner engagement states and their trajectories in discussion forums using two-cohort data from a graduate-level online course. We used HMM (Hidden Markov model) to discover the hidden longitudinal engagement clusters. Results showed two clusters in the first cohort and three clusters in the second cohort. Students tended to remain in a similar engagement state, and their participation patterns depended on the discussion dynamics likely influenced by class size.