Reproduce “VISAR: A Human-AI Argumentative Writing Assistant with Visual Programming and Rapid Draft Prototyping” 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
Zheng Zhang, Jie Gao, Ranjodh Singh Dhaliwal, Toby Jia-Jun Li
Researchers from the University of Notre Dame and Singapore University of Technology and Design developed VISAR, a human-AI assistant employing visual programming for outlining and rapid draft prototyping to support hierarchical and iterative planning in argumentative writing. The system helps users structure arguments and quickly evaluate plans through LLM-generated content, aiming to enhance the planning process.
Key insights
- Hierarchical ideation and 'argumentative sparks' are effective in assisting users with structured outlining, exploring ideas, and enhancing argument persuasiveness, despite some user concerns about perceived control.
- The synchronized visual and text interfaces provide complementary benefits, allowing users to better manage both the global logical structure and local content of their arguments.
- Rapid draft prototyping significantly helps users reflect on and improve their writing plans by making them tangible, although LLM-generated drafts require substantial human revision due to quality issues like repetition and poor transitions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.