Ever feel lost in jargon while watching technical videos - and stuck trying to find clear, timely help? Check out our work at #chi2025 on how real-time AI systems can bridge knowledge gaps without overwhelming users!
Yuhan Liu
@YuhanLiu@hci.social
- 150 Followers
- 86 Following
- 29 Posts
PhD candidate at PrincetonHCI Lab. Social Computing
This work was done when I was a summer intern at Accenture Labs, together with my awesome collaborators Aadit Shah, Jordan Ackerman, and Manaswi Saha.
Sadly, none of us will be in Japan — but we'll present online at the "Narratives and AI Support" session (https://programs.sigchi.org/chi/2025/program/session/197232) on Thursday, May 1, 9AM!
Read the full paper 👉 https://arxiv.org/abs/2503.00715
Sadly, none of us will be in Japan — but we'll present online at the "Narratives and AI Support" session (https://programs.sigchi.org/chi/2025/program/session/197232) on Thursday, May 1, 9AM!
Read the full paper 👉 https://arxiv.org/abs/2503.00715
We map out a design space for real-time LLM knowledge support systems, and discuss design implications for 1) balancing automation vs user control, 2) building trust in AI-generated info, 3) personalizing knowledge support, and 4) generalizing jargon explanation to other knowledge support
Findings? 📈
* StopGap provides real-time support and improves understanding without raising cognitive load.
* No "one-size-fits-all" knowledge representation formats: user preference depends on the context, the word, and their own habits and backgrounds.
* Participants want a balance between user-initiated and automated support, along with personalization and just-in-time assistance.
* StopGap provides real-time support and improves understanding without raising cognitive load.
* No "one-size-fits-all" knowledge representation formats: user preference depends on the context, the word, and their own habits and backgrounds.
* Participants want a balance between user-initiated and automated support, along with personalization and just-in-time assistance.
We built StopGap, a prototype that explains jargon in videos in various knowledge representation formats (definitions, metaphors, lists, and images), and tested it out through a design probe study (N=24)
sohyeon hwang📢 research call 📢
please boost! we're running a series of small group design workshops for community governance on fedi. if you are a user, admin, mod, dev, organizer - or have thought quite a bit about governance on fedi - and a legal adult in your locale, join us!
each workshop, scheduled based on peoples' availability, is:
🌐 2 hrs on Zoom
👥 w/ 6-10 people
💲 comes w/ $60 USD per person
details: https://dsmw.cs.princeton.edu/
sign up: https://princetonsurvey.az1.qualtrics.com/jfe/form/SV_e3SdG4y6wocMPie
questions: frictance@princeton.edu
Urgent: Looking for an external reviewer for a C&C 25 submission on GenAI and data sensemaking. The original reviewer accepted but ghosted, and the deadline is 3/15 AOE. Apologies for the super short notice!
If you have relevant experience and are interested, please DM me. Recommendations for potential reviewers are also greatly appreciated. Thanks!
Brian KruppPlease join us for the fourth ACM SIGCAS Works in Progress event of 2024 where Yuhan Liu will be giving a talk on "Understanding and Developing Decentralized Sociotechnical Systems"!
The talk will be on November 22nd, at 1pm EST
Registration is required for the event where you can register here:
https://forms.gle/2AUEVs9eHszGz6HTA
There will be a QA session after the talk and information on how to connect via Zoom will be sent the morning of the event.
Lei ZhangI'm recruiting 1-2 PhD students for Fall 2025 as I begin my journey as an assistant professor at the New Jersey Institute of Technology (NJIT) in January 2025. Let's explore the future of intelligent creation tools for AR/VR experiences. ✨ Details about my advising philosophy and research vision: https://raynez.art/join-us/.
🚀 Please help spread the word 🙏