Statistical Machine Translation
Enhanced Apertium System for Spanish–Asturian Translation
Pages
7
Time to read
21 mins
Publication
Language
English
Pages
7
Time to read
21 mins
Publication
Language
English
This document is a technical report detailing the enhancements made to the Spanish–Asturian Apertium translation system for the shared task focused on Low Resource Languages of Spain at the Ninth Conference on Machine Translation. The report outlines the background of the Eslema project, initiated by the University of Oviedo, which aims to develop a rule-based machine translator for the Spanish–Asturian language pair. It discusses the legislative context surrounding the Asturian language, recognizing it as a minority language with limited resources. The enhancements made to the Apertium system include grammatical, syntactic, and lexical improvements to increase translation accuracy. The report also describes the structure of the Apertium platform, including monolingual and bilingual packages, and the importance of various modules such as dictionaries, post-generators, and transfer rules. The methodology employed in the enhancements is explained, highlighting the involvement of linguists and the duration of the project.