arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.04020cs.CL
收藏

LLMAEL:大型语言模型是实体链接的良好上下文增强器

LLMAEL: Large Language Models are Good Context Augmenters for Entity Linking

Department of Computer Science and Technology, Beijing National Research Center for Information Science and Technology(计算机科学与技术系,北京信息科学国家研究中心);Tsinghua University(清华大学);Beijing University(北京大学)

Amy Xin, Yunjia Qi, Zijun Yao, Fangwei Zhu, Kaisheng Zeng, Xu Bin, Lei Hou, Juanzi Li

AI总结:

针对专门实体链接模型对长尾实体消解能力不足的问题,提出LLMAEL框架,利用无需微调的LLM生成实体描述作为上下文增强,在6个基准上取得最先进结果,准确率绝对提升8.9%。

摘要:

专门的实体链接(EL)模型擅长根据给定上下文将提及(mention)映射到知识库(KB)中的唯一实体。然而,由于训练数据有限,专门的实体链接模型难以消解长尾实体。与此同时,经过广泛预训练的大型语言模型(LLM)拥有关于不常见实体的更广泛知识。然而,由于缺乏专门的实体链接训练,大型语言模型经常无法生成准确的KB实体名称,限制了其在实体链接中的独立有效性。基于观察到大型语言模型更擅长上下文生成而非实体链接执行,我们引入了LLM增强实体链接(LLMAEL),这是第一个利用LLM数据增强来增强专门实体链接模型的框架。LLMAEL利用现成的、无需微调的LLM作为上下文增强器,生成实体描述作为专门实体链接模型的额外输入。实验表明,LLMAEL在6个广泛采用的实体链接基准上取得了新的最先进结果:与先前将无需微调的LLM集成到实体链接中的方法相比,LLMAEL在实体链接准确率上取得了8.9%的绝对提升。我们发布了我们的代码和数据集。

英文摘要:

Specialized entity linking (EL) models are well-trained at mapping mentions to unique knowledge base (KB) entities according to a given context. However, specialized EL models struggle to disambiguate long-tail entities due to their limited training data. Meanwhile, extensively pre-trained large language models (LLMs) possess broader knowledge of uncommon entities. Yet, with a lack of specialized EL training, LLMs frequently fail to generate accurate KB entity names, limiting their standalone effectiveness in EL. With the observation that LLMs are more adept at context generation instead of EL execution, we introduce LLM-Augmented Entity Linking (LLMAEL), the first framework to enhance specialized EL models with LLM data augmentation. LLMAEL leverages off-the-shelf, tuning-free LLMs as context augmenters, generating entity descriptions to serve as additional input for specialized EL models. Experiments show that LLMAEL sets new state-of-the-art results across 6 widely adopted EL benchmarks: compared to prior methods that integrate tuning-free LLMs into EL, LLMAEL achieves an absolute 8.9% gain in EL accuracy. We release our code and datasets.

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。

↑