Culture_datasets
Collection
3 items • Updated
Datasets:
Dataset Viewer
The dataset viewer is not available for this dataset.
The maximum number of configs allowed is 4000, dataset has 4335 configs.
Error code: DatasetWithTooManyConfigsError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Common Crawl Regional Corpus
A multilingual web corpus with geographic region annotations, extracted from Common Crawl. Each document is labeled with a locale (language-region pair).
Quick Start
from datasets import load_dataset
# Load a specific locale
ds = load_dataset("Yusser/CC-LARD", "es-MX")
# Load all data
ds = load_dataset("Yusser/CC-LARD", "all")
# Access the data
for row in ds["train"]:
print(row["text"][:100], row["locale"], row["region_source"])
Dataset Statistics
- Total documents: 13,150,911
- Languages: 140
- Regions: 349
- Locales: 4334
Columns
| Column | Type | Description |
|---|---|---|
text |
string | Document text |
lang |
string | ISO 639-1 language code |
region |
string | ISO 3166-1 alpha-2 region code |
locale |
string | Combined lang-region (e.g., es-MX) |
url |
string | Source URL |
lang_confidence |
float | fastText LID confidence |
region_source |
string | Signal used: html_lang, http_header, tld, none |
warc_date |
string | Crawl timestamp |
text_length |
int | Character count |
title |
string | Page title |
text_hash |
string | SHA-256 of normalized text |
Available Locales
4334 locales across 140 languages.
License
Derived from Common Crawl data, released under CC-BY-4.0.
- Downloads last month
- 1,138