Différences entre les versions de « Tyers (2009) »
De Arbres
(Page créée avec « * Tyers, Francis M. 2009. 'Rule-based augmentation of training data for breton–french statistical machine translation', ''Proceedings of the 13th Conference of the Europ... ») |
|||
Ligne 1 : | Ligne 1 : | ||
* Tyers, Francis M. 2009. 'Rule-based augmentation of training data for breton–french statistical machine translation', ''Proceedings of the 13th Conference of the European Association for Machine Translation'', 213–218. [http://xixona.dlsi.ua.es/~fran/publications/eamt2009b.pdf texte]. | * Tyers, Francis M. 2009. 'Rule-based augmentation of training data for breton–french statistical machine translation', ''Proceedings of the 13th Conference of the European Association for Machine Translation'', 213–218. [http://xixona.dlsi.ua.es/~fran/publications/eamt2009b.pdf texte]. | ||
'''Abstract''': | |||
"One characteristic feature of Celtic languages is [[mutation]], i.e. the fact that the initial consonant of words may change according to the context. We provide a quick description of this linguistic phenomenon for Breton along with a formalization using finite state transducers. This approach allows an exact and compact description of mutations. The result can be used in various contexts, especially for spell checking and language teaching." | |||
[[Category:ouvrages de recherche|Categories]] | [[Category:ouvrages de recherche|Categories]] | ||
[[Category:TAL|Categories]] | [[Category:TAL|Categories]] |
Version du 3 novembre 2021 à 23:16
- Tyers, Francis M. 2009. 'Rule-based augmentation of training data for breton–french statistical machine translation', Proceedings of the 13th Conference of the European Association for Machine Translation, 213–218. texte.
Abstract: "One characteristic feature of Celtic languages is mutation, i.e. the fact that the initial consonant of words may change according to the context. We provide a quick description of this linguistic phenomenon for Breton along with a formalization using finite state transducers. This approach allows an exact and compact description of mutations. The result can be used in various contexts, especially for spell checking and language teaching."