arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.05694cs.CLcs.AIcs.LG

中文零样本端到端关系抽取:Gemini、LLaMA 与 ChatGPT 的比较研究

Zero-Shot End-to-End Relation Extraction in Chinese: A Comparative Study of Gemini, LLaMA and ChatGPT

University of Amsterdam(阿姆斯特丹大学);Johns Hopkins University(约翰斯·霍普金斯大学);University of California San Diego(加利福尼亚大学圣迭戈分校);Simon Fraser University(西蒙菲莎大学);Northeastern University(东北大学)

Shaoshuai Du, Yiyi Tao, Yixian Shen, Hang Zhang, Yanxin Shen, Xinyu Qiu, Chuanqi Shi

AI总结:

本文比较了 ChatGPT、Gemini 和 LLaMA 在中文零样本端到端关系抽取中的表现,发现 ChatGPT 综合性能最佳,Gemini 推理最快,LLaMA 需进一步适配。

摘要:

本研究考察了多种大语言模型(LLMs)在中文零样本端到端关系抽取(RE)上的表现,该任务将实体识别和关系抽取整合在一起,且无需标注数据。尽管 LLMs 在 RE 上展现出潜力,但大多数先前工作集中于英语或假设实体已预先标注,导致其在中文 RE 中的有效性在很大程度上尚未被探索。为弥补这一空白,我们从准确率、效率和适应性三个方面评估了 ChatGPT、Gemini 和 LLaMA。ChatGPT 表现出最高的整体性能,在精确率和召回率之间取得平衡;而 Gemini 实现了最快的推理速度,使其适用于实时应用。LLaMA 在准确率和延迟方面均表现不佳,凸显了进一步适配的必要性。我们的研究结果为 LLMs 在零样本中文 RE 中的优势与局限提供了见解,并揭示了准确率与效率之间的权衡。本研究为未来旨在提升 LLMs 对中文 NLP 中复杂语言任务适应性的研究奠定了基础。

英文摘要:

This study investigates the performance of various large language models (LLMs) on zero-shot end-to-end relation extraction (RE) in Chinese, a task that integrates entity recognition and relation extraction without requiring annotated data. While LLMs show promise for RE, most prior work focuses on English or assumes pre-annotated entities, leaving their effectiveness in Chinese RE largely unexplored. To bridge this gap, we evaluate ChatGPT, Gemini, and LLaMA based on accuracy, efficiency, and adaptability. ChatGPT demonstrates the highest overall performance, balancing precision and recall, while Gemini achieves the fastest inference speed, making it suitable for real-time applications. LLaMA underperforms in both accuracy and latency, highlighting the need for further adaptation. Our findings provide insights into the strengths and limitations of LLMs for zero-shot Chinese RE, shedding light on trade-offs between accuracy and efficiency. This study serves as a foundation for future research aimed at improving LLM adaptability to complex linguistic tasks in Chinese NLP.

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。

↑