---
task_categories:
- feature-extraction
language:
- en
tags:
- Tags
- keywords
- generation
- news articles
- extraction
pretty_name: Tags Generated Dataset
size_categories:
- 1K<n<10K
---
<p align="center"><h1>🧠 Tags Generated Dataset</h1></p>

# Dataset Card for Tags Generated Dataset

- **Curated by:** Zineb MEFTAH  
- **Language(s) (NLP):** English  

---

## 📝 Dataset Summary

The Keyword Extraction Dataset includes 2,000 samples pairing extracted keywords with their corresponding full-text news articles. It is optimized for tasks such as keyword extraction, text analysis, and model fine-tuning.

- **Structure:**
  - 2,000 samples of keywords paired with full diversed themes news articles.
  - Column 1: Extracted keywords.
  - Column 2: Full text of the associated news article
- **Source:** Derived from the [BBC News dataset](https://www.kaggle.com/datasets/gpreda/bbc-news) available on Kaggle. The Kaggle dataset served as a foundational seed for iterative refinement.  

For more details on the dataset's creation and methodology, refer to the accompanying scientific [article](https://huggingface.co/blog/zino36/tags-generation-dataset-article).

---

## 🔍 Key Features

- **Comprehensive Coverage:** Multi-domain support for various themes tags generation.  
- **Optimized for AI Models:** Preprocessed and ready for fine-tuning.  

---

## 🌐 Licensing and Access

- **License:** Creative Commons Attribution 4.0 International (CC BY 4.0)  

---

## 📚 Citation

If you use this dataset, please cite:  

  - Author = Zineb Meftah
  - Title = Keyword Extraction Dataset
  - Year = 2024
  - Publisher = Hugging Face
  - License = CC BY 4.0