metadata

license: cc0-1.0
language:
  - se
tags:
  - corpus
  - saami
  - sami
size_categories:
  - 10M<n<100M

Northern Sámi Web Corpus

This open corpus is a result of a web crawl through admissible web pages in Northern Sámi. The crawl was seeded from the external links of the Sámi Wikipedia and continued with a breadth-first search through webpages that were identified as Northern Sámi using GlotLID and that allowed crawling according to their Robots Exclusion Protocol (robots.txt). The raw HTML documents were converted into natural text using Trafilatura. Finally, the corpus has been fuzzy-deduplicated on the document level. Our license does not impose any additional constraints on the contents of this corpus.