---
dataset_info:
  features:
  - name: url
    dtype: string
  - name: lang
    dtype: string
  - name: original_text
    dtype: string
  - name: tables
    dtype: string
  - name: processed_text
    dtype: string
  - name: english_content
    dtype: string
  - name: english_table
    dtype: string
  splits:
  - name: train
    num_bytes: 99021435
    num_examples: 12167
  download_size: 41056540
  dataset_size: 99021435
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: mit
task_categories:
- feature-extraction
---
# Dataset Card for `cookie-policy-corpus`

## Dataset Summary

The `cookie-policy-corpus` is a dataset that contains structured information from cookie policy web pages collected from real-world websites. Each entry includes the original webpage URL, raw policy content in its native language, extracted structured data (if available), and translated or processed content in both Vietnamese and English. The dataset supports research on cookie compliance detection, multilingual legal text processing, and translation tasks involving privacy-related content.

## Features

Each row in the dataset contains the following fields:

* **url** (`string`): The source URL of the webpage containing the cookie policy.
* **lang** (`string`): Language code of the original content (e.g., `es` for Spanish, `fr` for French).
* **original\_text** (`string`): Raw text extracted directly from the web page.
* **tables** (`list`): A list of any HTML tables parsed from the cookie policy page (if available).
* **processed\_text** (`string`): The original text processed or translated into Vietnamese (`vi`), mainly for normalization and semantic alignment.
* **english\_content** (`string`): A human-translated or machine-translated English version of the cookie policy text.
* **english\_table** (`list`): Extracted and translated cookie table(s) from the original page, if any are present.

## Languages
* Multilingual (source): Spanish (`es`), French (`fr`), Italian (`it`), etc.
* English (target standardization)

## Use Cases

This dataset can be used for:
* ✅ **Cookie Policy Compliance Analysis**: Analyzing how actual cookie usage aligns with declared policies across different countries and languages.
* 🌍 **Cross-lingual NLP**: Training or evaluating multilingual models for policy classification, translation, or legal reasoning tasks.
* 📚 **Legal Text Mining**: Extracting structured legal concepts (e.g., consent, third-party usage, retention) from natural language text.
* 🤖 **Multilingual Model Training**: Fine-tuning LLMs or translation models on privacy-related corpora.

## Example

```json
{
  "url": "https://tortugascw.com/privacy-cookies-policy/",
  "lang": "es",
  "original_text": "Links Privacy  Cookies Policy En el artículo 22.2 de la Ley 34[PHONE], de 11 de julio...",
  "tables": [],
  "processed_text": "Liên kết Chính sách về quyền riêng tư  cookie Trong Điều 22.2 của Luật 34...",
  "english_content": "Links Privacy  Cookies Policy In article 22.2 of Law 34[PHONE], of July 11...",
  "english_table": []
}
```

## Citation

If you use this dataset, please cite:

## License

This dataset is intended for research and educational purposes only. Content originates from publicly available web pages, and all rights belong to the respective site owners. Please verify terms of use if deploying for commercial purposes.