构建首个含空间标注的文档问答数据集，助力大模型理解图文内容。
BoundingDocs: a Unified Dataset for Document Question Answering with Spatial Annotations
- 将信息抽取等任务转为问答形式，适配大模型训练与评估
- 提供所有文档的OCR文本及答案在图像中的精确位置框
- 验证带空间信息的提示词对模型理解能力的提升效果
我们提出一个统一的文档问答（Document QA）数据集，整合了多个与文档AI和视觉丰富文档理解（VRDU）相关的公开数据集。主要贡献有二：一方面，将现有文档AI任务（如信息抽取）重构为问答任务，使其适合用于大语言模型的训练与评估；另一方面，公开所有文档的OCR文本，并标注答案在文档图像中的精确位置（边界框）。利用该数据集，我们研究了不同提示技术（可能包含边界框信息）对开源模型性能的影响，识别出最有效的文档理解方法。
原文摘要 · Abstract (English)
We present a unified dataset for document Question-Answering (QA), which is obtained combining several public datasets related to Document AI and visually rich document understanding (VRDU). Our main contribution is twofold: on the one hand we reformulate existing Document AI tasks, such as Information Extraction (IE), into a Question-Answering task, making it a suitable resource for training and evaluating Large Language Models; on the other hand, we release the OCR of all the documents and include the exact position of the answer to be found in the document image as a bounding box. Using this dataset, we explore the impact of different prompting techniques (that might include bounding box information) on the performance of open-weight models, identifying the most effective approaches for document comprehension.
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。