Traducción automática basada en caracteres y redes neuronales

Out of vocabulary words are still an open problem in translation systems. The aim of this work is trying to solve the problem of a limited vocabulary using translation levels smaller than words (e.g: characters). We want to take advantage of characteristics like word compounding, morphological and s...

Descripción completa

Detalles Bibliográficos
Autor: Larriba Flor, Antonio Manuel
Tipo de recurso: tesis de maestría
Fecha de publicación:2017
País:España
Institución:Universitat Politècnica de València (UPV)
Repositorio:RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Idioma:español
OAI Identifier:oai:riunet.upv.es:10251/89965
Acceso en línea:https://riunet.upv.es/handle/10251/89965
Access Level:acceso abierto
Palabra clave:Character-based Translation Models
Neural Machine Translation
Machine Translation
Traducción Automática
Traducción Automática basada en Redes Neuronales
Modelos de Traducción basados en Caracteres
LENGUAJES Y SISTEMAS INFORMATICOS
Máster Universitario en Inteligencia Artificial, Reconocimiento de Formas e Imagen Digital-Màster Universitari en Intel·ligència Artificial, Reconeixement de Formes i Imatge Digital
Descripción
Sumario:Out of vocabulary words are still an open problem in translation systems. The aim of this work is trying to solve the problem of a limited vocabulary using translation levels smaller than words (e.g: characters). We want to take advantage of characteristics like word compounding, morphological and semantical properties to try to translate unknown words from sub-word units already known. This way the translation model is not only able to deal with unseen words but also it will be capable of generating new words not seen during the training phase.