Abstract
We introduce CoDe-KG, an open-source, end-to-end pipeline for extracting sentence-level knowledge graphs by combining robust coreference resolution with syntactic sentence decomposition. Using our model, we contribute a dataset of over 150,000 knowledge triples, which is open source. We also contribute a training corpus of 7248 rows for sentence complexity, 190 rows of gold human annotations for co-reference resolution using open source lung-cancer abstracts from PubMed, 900 rows of gold human annotations for sentence conversion policies, and 398 triples of gold human annotations. We systematically select optimal prompt-model pairs across five complexity categories, showing that hybrid chain-of-thought and few-shot prompting yields up to 99.8% exact-match accuracy on sentence simplification. On relation extraction (RE), our pipeline achieves 65.8% macro-F1 on REBEL, an 8-point gain over the prior state of the art, and 75.7% micro-F1 on WebNLG2, while matching or exceeding performance on Wiki-NRE and CaRB. Ablation studies demonstrate that integrating coreference and decomposition increases recall on rare relations by over 20%. Code and dataset are available at this https URL
Abstract (translated)
我们介绍了CoDe-KG，这是一个开源的、端到端的流水线，用于结合稳健的共指消解与句法句子分解来提取句子级别的知识图谱。使用我们的模型，我们贡献了一个包含超过150,000条知识三元组的数据集，并且该数据集是开放源代码的。我们也提供了7248行关于句子复杂度的训练语料库、用PubMed上的开源肺癌摘要进行共指消解的人工标注黄金标准注释190行，转换策略的人工标注黄金标准注释900行以及398条人工标注的黄金标准三元组。我们系统地选择五个复杂性类别中的最优提示-模型对，结果表明混合链式思维和少量样本提示可以将句子简化的确切匹配准确率提高到最高达99.8%。 在关系抽取（RE）方面，我们的流水线在REBEL上达到了65.8%的宏观F1分数，比之前的最佳性能提高了8个百分点，并且在WebNLG2上的微平均F1得分为75.7%，同时在Wiki-NRE和CaRB数据集上的表现也与最佳水平持平或超越。消融实验表明，将共指消解和句子分解集成后，在罕见关系的召回率上可以提高超过20%。 代码和数据集可以在[提供的链接]中获取。
URL
https://arxiv.org/abs/2509.17289
