用真实分析笔记构建图表问答数据集，更贴近实际决策场景。
Chart Question Answering from Real-World Analytical Narratives
- 从可视化笔记中提取真实多视图图表与自然语言问题
- GPT-4.1在该数据集上准确率仅69.3%，显示现有模型仍不充分
- 适合研究真实世界图表理解与人机分析协作的学者
我们提出一个新图表问答（CQA）数据集，源自可视化笔记本。该数据集包含真实世界、多视图图表及基于分析叙事的自然语言问题。与以往基准不同，本数据反映生态有效推理流程。对先进多模态大模型的基准测试显示显著性能差距，GPT-4.1准确率为69.3%，凸显此更真实CQA设置的挑战。
原文摘要 · Abstract (English)
We present a new dataset for chart question answering (CQA) constructed from visualization notebooks. The dataset features real-world, multi-view charts paired with natural language questions grounded in analytical narratives. Unlike prior benchmarks, our data reflects ecologically valid reasoning workflows. Benchmarking state-of-the-art multimodal large language models reveals a significant performance gap, with GPT-4.1 achieving an accuracy of 69.3%, underscoring the challenges posed by this more authentic CQA setting.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。