Khondo: A Multimodal Benchmark for Document Packet Splitting of Bangla Forms
Khondo: A Multimodal Benchmark for Document Packet Splitting of Bangla Forms
Khondo：孟加拉语表单文档包拆分的基准测试
Abstract: Document packets, multiple documents concatenated into a single file, are common in government and administrative workflows, yet splitting them into their constituent documents is difficult, especially for low-resource languages.
摘要： 文档包（将多个文档合并为一个文件）在政府和行政工作流程中非常常见，但将其拆分为各个组成文档却十分困难，尤其是在低资源语言领域。
We introduce Khondo (Bangla for split/segment), the first benchmark for document packet splitting on Bangladeshi government forms. Unlike prior English and OCR-text-based datasets, Khondo is bilingual (Bangla—English) and vision-native; where models operate directly on page images.
我们推出了 Khondo（孟加拉语意为“拆分/分割”），这是首个针对孟加拉国政府表单进行文档包拆分的基准测试。与以往基于英语和 OCR 文本的数据集不同，Khondo 是双语（孟加拉语-英语）且原生视觉化的；模型直接在页面图像上进行操作。
It spans five concatenation schemes, from sequential to fully shuffled, across 14 administrative domains, with ground-truth boundaries, domain types, and page order. Zero-shot evaluation of MLLMs shows they cluster pages into their source documents fairly well but struggle in restoring the original page order once shuffled.
它涵盖了从顺序排列到完全打乱的五种拼接方案，涉及 14 个行政领域，并提供了真实边界、领域类型和页面顺序。对多模态大语言模型（MLLM）的零样本评估表明，它们能较好地将页面聚类到各自的源文档中，但在打乱顺序后，难以恢复原始页面顺序。
To isolate what drives this difficulty, we run two controlled analyses, varying the prompt instruction and then the packet language. Both primarily affect ordering rather than clustering: (a) explicit page-order instructions are necessary but insufficient, and (b) English packets are ordered more reliably than Bangla, making page arrangement the dominant challenge and language a secondary but consistent factor.
为了找出导致这一困难的原因，我们进行了两项对照分析，分别改变提示指令和文档包语言。这两者主要影响排序而非聚类：(a) 明确的页面顺序指令是必要的，但还不够；(b) 英语文档包的排序比孟加拉语更可靠，这使得页面排列成为主要挑战，而语言则是一个次要但持续存在的影响因素。
Khondo establishes page-order reconstruction as a key open problem in vision-based, low-resource document understanding, and provides a controlled benchmark for measuring progress toward solving it. Our dataset and code is available at this https URL.
Khondo 将页面顺序重构确立为基于视觉的低资源文档理解中的一个关键开放问题，并为衡量解决该问题的进展提供了一个受控的基准测试。我们的数据集和代码可在该链接获取。