Modern Standard Arabic NER corpus
收藏arXiv2023-04-07 更新2024-08-06 收录
下载链接：
http://arxiv.org/abs/2304.03399v1
下载链接
链接失效反馈官方服务：
资源简介：
Modern Standard Arabic NER corpus是由ITMO大学创建的一个针对阿拉伯语的命名实体识别数据集，包含超过36,000条记录，覆盖地理、历史、医学、体育、技术、新闻和烹饪七个领域。该数据集采用BIOES格式标注，支持处理嵌套命名实体，并被用于训练LSTM和GRU模型以提高阿拉伯语命名实体识别的准确性。数据集的创建旨在解决阿拉伯语领域内结构化数据不足的问题，并推动自然语言处理技术在该语言中的应用。
The Modern Standard Arabic NER corpus is a named entity recognition dataset for Arabic developed by ITMO University. It contains over 36,000 records spanning seven domains: geographical, historical, medical, sports, technical, news, and culinary. Annotated in the BIOES format, this corpus supports the handling of nested named entities and has been utilized to train LSTM and GRU models to enhance the accuracy of Arabic named entity recognition. The creation of this corpus aims to address the shortage of structured data within the Arabic language domain and promote the application of natural language processing technologies in Arabic.
提供机构：
ITMO大学
创建时间：
2023-04-07
