Interlingua based neural machine translation

We propose a machine translation architecture based on autoencoders and a shared interlingua representation that produce comparable results to state of the art systems. Also we define evaluation and visualization strategies as metrics of the performance of the architecture.

Bibliographic Details
Author: Escolano Peinado, Carlos|||0000-0001-6657-673X
Format: master thesis
Publication Date:2018
Country:España
Institution:Universitat Politècnica de Catalunya (UPC)
Repository:UPCommons. Portal del coneixement obert de la UPC
Language:English
OAI Identifier:oai:upcommons.upc.edu:2117/121672
Online Access:https://hdl.handle.net/2117/121672
Access Level:Open access
Keyword:Neural networks (Computer science)
Machine learning
deep learning
autoencoders
machine translation
Xarxes neuronals (Informàtica)
Aprenentatge automàtic
Àrees temàtiques de la UPC::Informàtica
Description
Summary:We propose a machine translation architecture based on autoencoders and a shared interlingua representation that produce comparable results to state of the art systems. Also we define evaluation and visualization strategies as metrics of the performance of the architecture.