Jerry Liu推文
PDF解析的乐趣：企业文档的无限多样性
PDF解析很有趣，因为企业文档种类繁多。对于每种文档，都需要精确的边界框、置信度和领域特定标注，以便为下游代理提供丰富元数据。以表单为例，除了输出为markdown，我们检测每个注释、字段、复选框和部分，从而无需单独的LLM提取步骤即可获得结构化信息，并附带来源引用。这很难，但值得优化。查看LlamaParse！
译文
PDF解析之所以有趣，是因为企业文档的形态无穷无尽 📑。对于每一类文档，背后都有一长串工作要做：构建更精确的边界框、置信度分数，以及领域特定的标注，从而为任何下游智能体提供丰富的元数据，而不必让它从头再来一遍。以表单为例。除了简单地将其输出为markdown之外，我们还投入精力去检测每一个批注、字段、复选框和区块。这样一来，你就能直接获得结构化信息，判断表单是否已填写，而无需额外的LLM提取步骤。你还能免费获得来源引用！把这件事做好并不容易。✅ 你可以做的优化深不见底，包括为每种文档类型提取批注。✅ 要正确渲染图表、手写内容等视觉格式并将其转化为数字化信息，难度很大。你越是跳过这一步，就给下游智能体制造越多的工作量。✅ 精确的边界框是任何类型文档精确引用的必要条件。你可以激进地调优模型和工具链，使得在任意文档子类型上的准确率/成本比前沿模型更具竞争力。无论你是在解析表单（参见“处理选项”中的enriched forms选项）还是其他类型的文档，都欢迎试试LlamaParse！https://cloud.llamaindex.ai/
PDF parsing is fun because there's an infinite variety of enteprise documents 📑. For each document category, there's a long tail of work to build more precise bounding boxes, confidence scores, and domain-specific annotations so that you provide any downstream agent rich metadata without it having to reinvent this from scratch. Take forms for example. Besides simply outputting it into markdown, we put in the work to detect every annotation, field, checkbox, and section. That way you immediately get structured information as to whether a form is filled without a separate LLM extraction step. You also get source citations for free! Doing this well is hard. ✅ There's a rabbit-hole of optimizations you can do, including extracting annotations for every document type. ✅ It's hard to properly render visual formats like charts, handwriting into digitalized information. The more you skip this step, the more work you're creating for any downstream agent. ✅ Precise bounding boxes are a necessity for precise citations on any type of document. You can aggressively tune the model + harness so that the accuracy/cost on any document subtype is much more competitive than the frontier models. Whether you're parsing forms (see the enriched forms option in "processing options") or any other type of doc, check out LlamaParse ! https://cloud.llamaindex.ai/
