Beyond the Page: Enriching Academic Paper Reading with Social Media Discussions
summary
Researchers build SURF, an interface that pulls social-media posts by DOI/X/Google, parses PDFs, uses LLMs (prompt-tuned with DSPy) to map comments to paper sections and classify them into 8 thread types and tutorials. A formative probe (N=8) guides design; a within-subjects study (N=18) shows significantly deeper comprehension, higher self-efficacy and lower cognitive load when SURF surfaces tutorial threads, color-coded signifiers, hover summaries, subthread extraction and quality-ranked feeds while hiding trivial “congrats” noise.
AI-written summary · about
chapters (14)
- 0:00Introduction and motivation
- 1:28Formative study findings
- 2:10Surf interface design
- 3:33Paper overview feature
- 4:15Contextualizing discussions
- 5:17Highlighting insightful threads
- 5:58Data pipeline and automation
- 6:19Usability study results
- 7:20Benefits for peer review
- 8:45Broader implications
- 9:48Potential risks and biases
- 10:10Open questions and future work
- 11:12Conclusion and thanks
- 11:39Q&A session
related talks
24mWhat and Why, How to Leverage Data Science and Research
01h 00mStanding on the Shoulders of Giants
09mHow to Support Users in Understanding Intelligent Systems? Structuring the Discussion
33mTired of the state of online conversations?
10mFrom Philosophy to Interfaces: an Explanatory Method and a Tool Inspired by Achinstein's Theory ...
45mThe Content Design of Civil Discourse
22mSocial Computing-Driven Activism in Youth Empowerment Organizations: Challenges and Opportunities
07mDLD Fellowship