A Free/Open-Source Morphological Analyser and Generator for Sakha

Sardana Ivanova, Jonathan Washington, Francis M. Tyers

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Abstract

We present, to our knowledge, the first ever published morphological analyser and generator for Sakha, a marginalised language of Siberia. The transducer, developed using HFST, has coverage of solidly above 90%, and high precision. In the development of the analyser, we have expanded linguistic knowledge about Sakha, and developed strategies for complex grammatical patterns. The transducer is already being used in downstream tasks, including computer assisted language learning applications for linguistic maintenance and computational linguistic shared tasks.
Original languageEnglish
Title of host publicationLREC 2022, THIRTEEN INTERNATIONAL CONFERENCE ON LANGUAGE RESOURCES AND EVALUATION : LREC 2022 Conference Proceedings
Number of pages6
PublisherEuropean Languages Resources Association (ELRA)
Publication dateJun 2022
Pages5137-5142
ISBN (Electronic)979-10-95546-72-6
Publication statusPublished - Jun 2022
MoE publication typeA4 Article in conference proceedings
EventLanguage Resources and Evaluation Conference - Marseille, France
Duration: 21 Jun 202223 Jun 2022
Conference number: 13
https://lrec2022.lrec-conf.org/en/

Fields of Science

  • 113 Computer and information sciences

Cite this