electricsheepafrica/africa-attack-of-nation-media

该数据集名为Attack of Nation Media，由Africa Uncensored发布，数据来源于OpenAfrica。数据集包含20,000行数据，分为16,000行的训练集和4,000行的测试集，涵盖39列变量，包括地理、时间、人口统计、标识符/元数据等多个方面。数据集主要关注非洲多个国家的社交媒体数据，特别是使用#TaifaLeoFakeNews标签的X（原Twitter）帖子。数据集经过Electric Sheep Africa处理，转换为适合机器学习的Parquet格式。数据集存在一些限制，如数据未独立验证、部分列缺失值较多等。

The dataset is named Attack of Nation Media, published by Africa Uncensored and sourced from OpenAfrica. It contains 20,000 rows of data, divided into a training set of 16,000 rows and a test set of 4,000 rows, covering 39 columns of variables including geographic, temporal, demographic, identifier/metadata, and other aspects. The dataset primarily focuses on social media data from multiple African countries, particularly X (formerly Twitter) posts using the #TaifaLeoFakeNews hashtag. The dataset has been processed by Electric Sheep Africa and converted into a machine-learning-ready Parquet format. There are some limitations to the dataset, such as unverified data and a high percentage of missing values in some columns.