arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.01838cs.CL
收藏

AXE:低成本跨领域网络结构化信息提取

AXE: Low-Cost Cross-Domain Web Structured Information Extraction

Faculty of Computers & Artificial Intelligence - Cairo University(开罗大学计算机与人工智能学院);Ejada Systems;Fawry Integrated Systems;Microsoft(微软)

Abdelrahman Mansour, Khaled W. Alshaer, Moataz Elsaban

AI总结:

AXE通过将HTML DOM视为需修剪的树结构,利用专用修剪机制提取高密度上下文,使小型模型生成精确结构化输出,并通过GXR确保提取可追溯性,实现跨领域信息提取的高性能与低成本。

摘要:

提取网络结构化数据往往是在手动启发式方法的脆弱性和大型语言模型的高成本之间进行权衡。我们介绍了AXE(自适应X路径提取器),一种重新思考这一过程的流程,通过将HTML DOM视为需要修剪的树结构,而不是仅仅是一堵要阅读的文本墙。AXE使用专门的“修剪”机制去除冗余和无关节点,留下浓缩的高密度上下文,使小型0.6B LLM能够生成精确的结构化输出。为了保持模型诚实,我们实现了基于XPath的地面真相解析(GXR),确保每次提取都能物理上追溯到源节点。尽管其低开销,AXE在零样本性能上达到最先进的水平,在SWDE数据集上以88.1%的F1分数超越了几个更大、完全训练过的替代方案。通过发布我们的专用适配器,我们旨在提供一种实用且成本效益高的大规模网络信息提取途径。我们的代码和适配器在https://github.com/abdo-Mansour/axetract上公开可用。

英文摘要:

Extracting structured data from the web is often a trade-off between the brittle nature of manual heuristics and the prohibitive cost of Large Language Models. We introduce AXE (Adaptive X-Path Extractor), a pipeline that rethinks this process by treating the HTML DOM as a tree that needs pruning rather than just a wall of text to be read. AXE uses a specialized "pruning" mechanism to strip away boilerplate and irrelevant nodes, leaving behind a distilled, high-density context that allows a tiny 0.6B LLM to generate precise, structured outputs. To keep the model honest, we implement Grounded XPath Resolution (GXR), ensuring every extraction is physically traceable to a source node. Despite its low footprint, AXE achieves state-of-the-art zero-shot performance, outperforming several much larger, fully-trained alternatives with an F1 score of 88.1% on the SWDE dataset. By releasing our specialized adaptors, we aim to provide a practical, cost-effective path for large-scale web information extraction. Our code and adaptors are publicly available at https://github.com/abdo-Mansour/axetract.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑