Reproduce “VizLinter: A Linter and Fixer Framework for Data Visualization” 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
Qing Chen, Fuling Sun, Xinyue Xu, Zui Chen, Jiazhe Wang, Nan Cao
VizLinter is a framework that detects fundamental errors in data visualization specifications and automatically provides and applies optimal corrections to resolve them. In a user study, the system achieved a 90% acceptance rate for its automated fixing suggestions and helped prevent the introduction of new errors during visualization development.
Key insights
- Answer Set Programming (ASP) provides a robust and easily extensible declarative framework for defining and updating visualization validation rules.
- Formulating the multi-error correction task as a Binary Integer Programming problem enables the system to find an optimal set of non-redundant actions that maximize problem resolution while minimizing cost.
- Automating both error detection and correction not only streamlines the visualization development process but also serves as an effective pedagogical tool for learning visualization principles.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.