Semi-automatic UNL Dictionary Generation using WordNet.PT


Catarina Ribeiro, Ricardo Santos, Rui Pedro Chaves, Palmira Marrafa

Universidade de Lisboa, CLUL, CLG - Computation of Lexical and Grammatical Knowledge Research Group, Av. Prof. Gama Pinto, 2, 1649-003 Lisboa, Portugal




The increase of Internet users all over the world and the subsequent growth of available multilingual information on the Web have brought new challenges to machine translation systems. The Universal Networking Language (UNL) is a meta-language developed for conveying linguistic expressions in order to encode websites information into a standard representation. In order to integrate Portuguese into this platform it is necessary to develop both a dictionary and a grammar. This paper focuses on the development of a PT-UNL dictionary using the WordNet.PT knowledge base. UNL makes use of lexical-semantic relations that have some correspondence in WordNet.PT. Since building the dictionary is a very time-consuming task, a semi-automatic process to generate it has been developed and is described here. The reuse of manually built lexical resources is of great relevance both for the economy of the project and for the reliability of the results.


Universal concepts/words, Linguistic Knowledge Base, Lexical-semantic relations, Inheritance.

Language(s) Multilingual
Full Paper