GLITTER: An AI-assisted Platform for Material-Grounded Asynchronous Discussion in Flipped Learning
summary
GLITTER is a web platform that links readings (left pane) to movable discussion blocks (right pane) and uses LLM-based features: semantic post navigation surfaces conceptually related posts, a one-click summarizer creates bullet summaries, and a drag-and-drop conceptual blending tool extracts three aspects from two posts to generate blended prompts plus three evidence snippets from the readings. It produces personalized interactive reports that visualize reading paths and gaps. Lab study (n = 12) and course deployment show increased engagement, new ideas, stronger reflection, and better preparedness.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 0:24Flipped learning challenges
- 1:08Cognitive challenges in discussions
- 2:12Design goals for GLIDER
- 2:55GLIDER features overview
- 3:58Public mode and navigation
- 4:40Summarization and blending
- 6:44Metacognitive reflection reports
- 7:26User study results
- 8:48Key takeaways and future work
- 9:29Q&A session
related talks
18mLeveraging LLMs for student feedback in introductory data science courses
01h 23mEmpowering Learning: IT, Education, and Ethical Considerations
01h 04mDesigning Future-Ready Classrooms: Leveraging AI & OER to Support Diverse Learners
01h 33mRAG How To Talk To Your Data
19mHacking Productivity with LLMs
48mBeyond the Prompt: Evaluating, Testing, and Securing LLM Applications
42mTo RAG or not to RAG
23mAn Interactive Learning Environment for Program Design