CITATION.cff

authors:
  - family-names: Wiechetek
    given-names: Linda
    orcid: "https://orcid.org/0000-0002-5171-0841"
  - family-names: Unhammer
    given-names: Kevin Brubeck
    orcid: "https://orcid.org/0000-0002-2883-1899"
  - family-names: Moshagen
    given-names: Sjur Nørstebø
    orcid: "https://orcid.org/0000-0003-3771-9521"
cff-version: 1.2.0
identifiers:
  - description: Workshop on the Use of Computational Methods in the Study of Endangered Languages
    type: url
    value: https://computel-workshop.org/wp-content/uploads/2019/02/CEL3_book_papers_draft.pdf#page=58
keywords:
  - Sámi
  - Saami
  - North Saami
  - proofing
  - grammar checking
  - grammar checker
  - spellcheck
  - tokenisation
  - FST
  - HFST
message: If you use this software, please cite it using these metadata.
repository-code: "https://github.com/divvun/libdivvun"
title: Divvun gramcheck
version: 0.3.10
preferred-citation:
  authors:
    - family-names: Wiechetek
      given-names: Linda
    - family-names: Unhammer
      given-names: Kevin Brubeck
    - family-names: Moshagen
      given-names: Sjur Nørstebø
  title: "Seeing more than whitespace—Tokenisation and disambiguation in a North Sámi grammar checker"
  type: article
  year: 2019
  url: "https://computel-workshop.org/wp-content/uploads/2019/02/CEL3_book_papers_draft.pdf#page=58"
  abstract: "Communities of lesser resourced languages like North Sámi benefit from language tools such as spell checkers and grammar checkers to improve literacy. Accurate error feedback is dependent on well-tokenised input, but traditional tokenisation as shallow preprocessing is inadequate to solve the challenges of real-world language usage. We present an alternative where tokenisation remains ambiguous until we have linguistic context information available. This lets us accurately detect sentence boundaries, multiwords and compound error detection. We describe a North Sámi grammar checker with such a tokenisation system, and show the results of its evaluation."
license: GPL-3.0-or-later
url: https://github.com/divvun/libdivvun