Back to Gap Map Model components / Language-specific datasets

Northern Sámi web corpus

Language Technology Group, University of Oslo
open / Overall score: 1.7

The Northern Sámi web corpus is a crawl of about 33,000 web documents in Northern Sámi, seeded from the external links of the Sámi Wikipedia and continued breadth-first through pages that GlotLID identified as Northern Sámi and whose robots.txt allowed crawling. Text was extracted with Trafilatura and fuzzy-deduplicated document by document. The Language Technology Group at the University of Oslo publishes it.

Openness

5 high confidence
5.0
license
cc0-1.0
access
public (Hugging Face gated: false.)
dataset_card
present (Short card: crawl seeding, language filter, extraction and deduplication; no size statistics.)

The corpus is dedicated to the public domain under CC0 and downloads without a gate. The crawl kept only pages whose robots.txt allowed it, and the card says the license adds no constraints on the crawled content.

Adoption

1 high confidence
1.0

Hugging Face downloads of the dataset repository over the trailing 30 days. A download is a file fetch rather than a trained model, and copies mirrored elsewhere are not counted.

Capability

2 medium confidence
2.0

The corpus gives Northern Sámi a crawled web collection whose short card explains how the crawl was seeded, filtered and deduplicated, and TurkuNLP lists it among the training data of its Finnish ModernBERT models. Judged against what Northern Sámi has, an official language with under a million speakers and a small written web, its few tens of millions of tokens and missing paper put it one step below Korpus Malti.

Verified 2026-09-24