DV-World: Benchmarking Data Visualization Agents in Real-World Scenarios
2026-04-28 • Computation and Language
Computation and Language
AI summaryⓘ
The authors created DV-World, a set of 260 tasks to better test how well computer programs can help with making data visualizations in real work situations. These tasks cover working directly with spreadsheets, updating visualizations as data changes, and understanding unclear user requests. They also developed ways to check if the programs are accurate and make sense visually. Their tests showed current models still struggle with these tasks, scoring less than 50%. DV-World aims to help improve tools that assist with data visualization in practical settings.
data visualizationbenchmarkspreadsheet manipulationcross-platformintent alignmentuser simulatorMLLMnumerical precisionsemantic assessmententerprise workflows
Authors
Jinxiang Meng, Shaoping Huang, Fangyu Lei, Jingyu Guo, Haoxiang Liu, Jiahao Su, Sihan Wang, Yao Wang, Enrui Wang, Ye Yang, Hongze Chai, Jinming Lv, Anbang Yu, Huangjing Zhang, Yitong Zhang, Yiming Huang, Zeyao Ma, Shizhu He, Jun Zhao, Kang Liu
Abstract
Real-world data visualization (DV) requires native environmental grounding, cross-platform evolution, and proactive intent alignment. Yet, existing benchmarks often suffer from code-sandbox confinement, single-language creation-only tasks, and assumption of perfect intent. To bridge these gaps, we introduce DV-World, a benchmark of 260 tasks designed to evaluate DV agents across real-world professional lifecycles. DV-World spans three domains: DV-Sheet for native spreadsheet manipulation including chart and dashboard creation as well as diagnostic repair; DV-Evolution for adapting and restructuring reference visual artifacts to fit new data across diverse programming paradigms and DV-Interact for proactive intent alignment with a user simulator that mimics real-world ambiguous requirements. Our hybrid evaluation framework integrates Table-value Alignment for numerical precision and MLLM-as-a-Judge with rubrics for semantic-visual assessment. Experiments reveal that state-of-the-art models achieve less than 50% overall performance, exposing critical deficits in handling the complex challenges of real-world data visualization. DV-World provides a realistic testbed to steer development toward the versatile expertise required in enterprise workflows. Our data and code are available at \href{https://github.com/DA-Open/DV-World}{this project page}.