Publication - Parameter Optimisation of MT Systems Operating on Monolingual Corpora Employing a Genetic Algorithm
PROJECTS

Parameter Optimisation of MT Systems Operating on Monolingual Corpora Employing a Genetic Algorithm

Research Area:  
Other topics in Computer Science
    
Type:  
In Proceedings

 

Year: 2008
Authors: Sokratis Sofianopoulos; George Tambouratzis; George Carayannis
Editor: Marik V.; Bradshaw J.M.; Meyer J.; Gruver W.A.; Benda P.
Book title: Proceedings of the IEEE International Conference on Distributed Human-Machine Systems (DHMS’08)
Pages: 260-265
Address: Αθήνα, Ελλάδα
Organization: IEEE
Date: 9-12 Μαρτίου
ISBN: 978-80-01-04027-0
Abstract:
In this paper, a method is proposed for the automated optimisation of parameters used in a system performing automatic translation from a source language to a target language. This machine translation system is based on pattern recognition techniques together with extensive monolingual corpora in the target language to achievethe translation of text. The proposed optimisation method is based on applying a Genetic Algorithm to tune the parameters of the system towards the generation of high quality translations. The Genetic Algorithm is guided by an established metric of the machine translation quality.The experimental results reported indicate the effectiveness of this approach, since it results, for the development set of sentences, in translations that are of a substantially higher quality than these achieved by the same system when the parameters are fine-tuned manually, by specialised scientists.
[Bibtex]