2022 Fiscal Year Annual Research Report
Personalized Online Adaptive Learning System
| Project/Area Number |
20H01719
|
|---|
| Allocation Type | Single-year Grants |
|---|
| Research Institution | Tokyo Institute of Technology |
|---|
Principal Investigator |
クロス ジェフリーS 東京工業大学, 環境・社会理工学院, 教授 (90532044)
|
|---|
| Project Period (FY) |
2020-04-01 – 2025-03-31
|
| Keywords | dashboard / AI / personalized learning / machiine learning / online learning / metacognition / LMS / software |
|---|
| Outline of Annual Research Achievements |
The Personalized Online Adaptive Learning System (POALS) is a web-based add-on to a learning management system (LMS) developed to help learners succeed when taking online learning courses. It is made up of three components: 1) the Metacognitive Tutor to equip students with metacognitive skills needed for autonomous learning crucial to online learning, 2) the Adaptive Engine to help students manage the cognitive strain of having metacognitive tutoring alongside domain knowledge learning, and the 3) Analytics Dashboard based on student responses on the Metacognitive Tutor to give feedback to teachers on where learner interventions might be needed. In 2021, a variation of POALS' Metacognitive Tutor was added to AI Kaku, an AI-assisted writing tool also created at the PI's laboratory. In 2022, the focus had been on adaptive learning through capacity-building in gamification and virtual reality. Key takeaways from this research include which metacognitive skills can be learned independently and which ones require nudging from teachers, which algorithms can work best for knowledge tracing while considering metacognitive measures, and how text data from learner metacognitive reflections can be used for learning analytics. Research achievements for 2022 include one peer review journal paper published and 9 conference presentations for both domestic and international conferences.
|
| Current Status of Research Progress |
Current Status of Research Progress
1: Research has progressed more than it was originally planned.
Reason
We recently external support from a company that helped developed part of the integrated POALS dashboard for the instructor. As a result the research is a head of schedule.
|
| Strategy for Future Research Activity |
The POALS research plan for AY 2023 will focus on evaluating an Instructor Analytics Dashboard. The dashboard is needed to summarize the learner assessment data via graphs and natural language processed narratives that the instructor can use to feedback to the learner to encourage learning and course engagement. Large language models similar to those used by ChatGPT will be used for create feedback narratives to be displayed in the Analytics Dashboard. Aside from the NLP flair, the Analytics Dashboard is being upgraded to support drilldown analytics. The effectiveness of the drilldown analytics will be analyzed from the human-computer interaction (how usable is the system) and informatics (how useful is the information) lenses. The work on Analytics Dashboard may span up to two years. Elementary school teacher readiness in teaching programming from two years ago will be resumed. This complementary research is expected to contribute to the overall goal of POALS by shedding an understanding on teachers' capability to use ICT for instruction, which is the condition where POALS is situated at. In addition, research results which were obtained last year will be submitted as journal papers and submitted to international conferences.
|
