Reproduce “ITO: Real-time Browser Tab Orchestration Through Intent Detection” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Elvin Hu, Jasmine Sumpter, Takatoshi Yoshida, Mauricio Sousa, Kouta Minamizawa
Ito is a mixed-initiative system that orchestrates browser tabs in real time by continuously inferring user intent from implicit browsing signals and organizing them into dynamic "Flows". Evaluation revealed that this approach aligns with users' mental models, reduces tab management burden, and enhances task switching without compromising user control, showing increased user comfort over a two-week field study.
Key insights
- Continuous, real-time intent inference from implicit browsing signals effectively aligns browser organization with users' dynamic mental models, reducing cognitive load without requiring explicit user declarations.
- A mixed-initiative approach with continuous, additive operations and multi-scale correction mechanisms can reduce tab management "meta-activity" while crucially preserving users' sense of control and agency.
- Supporting the full lifecycle of organized tab groups, including proactive hibernation and intelligent resumption, is essential for maintaining continuity across interleaved and long-lived browsing tasks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.