CNER-UAV
收藏资源简介:
CNER-UAV是由香港城市大学计算机科学系创建的细粒度中文地址实体识别数据集,专为无人机配送系统中的地址解析任务设计。该数据集包含约12,000个标注样本,涵盖五种不同类别,数据来源于美团无人机配送系统,经过严格的数据清洗和去标识化处理。创建过程中,数据集通过人类专家和大型语言模型(如GPT-3.5和ChatGLM)进行标注,形成了三个子集。CNER-UAV主要用于解决无人机配送系统中地址解析的精确性和效率问题,是目前中国最全面、最新的地址数据集之一。
CNER-UAV is a fine-grained Chinese address entity recognition dataset developed by the Department of Computer Science, City University of Hong Kong, tailored specifically for address parsing tasks in unmanned aerial vehicle (UAV) delivery systems. The dataset contains approximately 12,000 annotated samples spanning five distinct categories, with its source data originating from the Meituan UAV delivery system, and has been subjected to rigorous data cleaning and de-identification processing. During its creation, the dataset was annotated by human experts and large language models (such as GPT-3.5 and ChatGLM), yielding three subsets. CNER-UAV is primarily intended to address the accuracy and efficiency issues of address parsing in UAV delivery systems, and is currently among the most comprehensive and up-to-date address datasets in China.

- 1Can LLM Substitute Human Labeling? A Case Study of Fine-grained Chinese Address Entity Recognition Dataset for UAV Delivery香港城市大学计算机科学系 · 2024年



