This project is funded by Programa de Promoción Xeral da Investigación, Consellaría de Innovación e Industria.
The objective of this project is to elaborate an automatic extractor of bilingual lexicons for Galician and Spanish. In addition, we aim at using those lexicons in order to improve and update the computational dictionaries underlying machine translation systems, such as Opentrad. The extraction method is based on non-parallel, comparable corpora. Special attention will be paid to the extraction of multi-word translation equivalents.
Sintax Information Extraction Dependency Grammar Comporable Corpora Bilingual Dictionaries Parsing