VALUABLY-net/iab-taxonomy-multilang-content

这个数据集包含用于IAB（互动广告局）分类的多语言文本数据。数据文件（train.csv，val.csv）是通过处理、清洗和重构来自Kaggle上原始发布的IAB-Taxonomy URL Content Dataset生成的。伴随的处理脚本和JSON映射文件提供了数据准备和模型训练工作流程的支持。数据集使用IAB第二层类别（共392个标签）进行多类分类。

This dataset contains multilingual text data for IAB (Interactive Advertising Bureau) taxonomy classification. The data files (`train.csv`, `val.csv`) were generated by processing, cleaning, and restructuring data file from the [IAB-Taxonomy URL Content Dataset](https://www.kaggle.com/datasets/ishikaaaaaathakur/iab-taxonomy-url-content-dataset) originally published on Kaggle. The accompanying processing scripts and JSON mapping files are provided to support the data preparation and model training workflow. The dataset uses IAB Tier 2 categories (392 labels total) for multi-class classification.