harem

HAREM是一个葡萄牙语语料库，专为命名实体识别任务设计。它包含约9.3万个单词，来源于129篇不同类型和语言风格的文本，并提供包含10个命名实体类别和仅包含5个类别的两种版本。该语料库的数据集划分遵循特定文献，分为训练集、验证集和测试集。数据以标准化格式提供，包含文本tokens和对应的NER标签，方便进行数据操作。

HAREM is a Portuguese-language corpus specifically designed for named entity recognition (NER) tasks. It contains approximately 93,000 words, sourced from 129 texts with diverse types and linguistic styles, and provides two variants: one with 10 named entity categories and the other with only 5 categories. The dataset splits of this corpus follow specific academic literature, and are divided into training, validation, and test sets. The data is provided in a standardized format, including text tokens and their corresponding NER tags, which facilitates data manipulation.