Poesiaren eskantsio automatikoa: bi hizkuntzen azterketa
No Thumbnail Available
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Lan honetan poesiaren eskantsioa, hau da, poemetako egitura erritmikoaren erauztea, burutzen duguautomatikoki. Horretarako hizkuntzaren prozesamenduko ohiko teknikak eta sare neuronaletan oinarri-tutakoak erabili ditugu. Esperimentazioa bi hizkuntzarekin egin dugu, ingelesarekin eta gaztelerarekin.Emaitzen arabera, sare neuronalekin lortu ditugu emaitza onenak, bi hizkuntzetan. Gainera, ondorioztadaiteke gaztelerazko poesia erregularragoa dela ingelesezkoa baino.Egungo erronka esperimentuokeuskarara aplikatzea da.
In this work we perform automatic scansion —the act of metrically representing a line of verse— ofpoetry. We employ typical Natural Language Processing techniques and go further by using current DeepLearning models. We have tested our systems in two languages; English and Spanish. We got our bestresults with the Deep Learning frameworks, both in English and Spanish. Furthermore, we can concludethat the scansion of poetry in English is more complex than in Spanish. We intend to apply these modelson Basque poetry.
In this work we perform automatic scansion —the act of metrically representing a line of verse— ofpoetry. We employ typical Natural Language Processing techniques and go further by using current DeepLearning models. We have tested our systems in two languages; English and Spanish. We got our bestresults with the Deep Learning frameworks, both in English and Spanish. Furthermore, we can concludethat the scansion of poetry in English is more complex than in Spanish. We intend to apply these modelson Basque poetry.
Description
Keywords
literatura, poesia, eskantsioa, ikasketa automatikoa, ingelesa, gaztelera, euskara, literature, poetry, scansion, machine learning, English, Spanish, Basque