PARQR: Automatic Post Suggestion in the Piazza Online Forum to Support Online Students
summary
PARQR augments Piazza with a real‑time recommender that parses a draft post, shows similar threads with metadata (follow-ups, age, week), and adds a TA dashboard of high‑priority unanswered items; deployed across five online master's courses, it surveys 74 students and interviews five TAs. Evaluation shows lower posting rates 0.291 vs. 0.506 posts per active student and fewer duplicates 17.8% vs. 25.6% duplicate posts, while users request clearer recommendation rationale and improved visual urgency cues.
AI-written summary · about
chapters (9)
- 0:00Start
- 0:46Motivation
- 2:14Related Work
- 3:48Post Recommendations View
- 4:18Dashboard View
- 4:43Methods for Data Collection
- 5:13Research Timeline
- 9:18Summary of Findings
- 9:48Research Contribution
related talks
18mLeveraging LLMs for student feedback in introductory data science courses
11mPosterMate: Audience-driven Collaborative Persona Agents for Poster Design
12mGLITTER: An AI-assisted Platform for Material-Grounded Asynchronous Discussion in Flipped Learning
12mPOET: Supporting Prompting Creativity and Personalization with Automated Expansion of Text-to-Ima...
42mPrompt. Build. Repeat. – The Vibe Coding Mindset
24mAI: Empowering Students, Not Replacing Universities
14mUnconference: Self-optimizing deep research agents
39mChat with Me and Get the PromQL You Want