RuDriCo2 - a faster disambiguator and segmentation modifier
Cláudio Diniz, Nuno Mamede, João Pereira, Alves Redol · 2010
Currently, L 2 F 's NLP chain has a bottleneck. Module Ru- DriCo (Rule Driven Converter) is substantially slower than the remaining modules of the chain. RuDriCo is a rule-based morphological disambigua- tor with the possibility to change segmentation (join or split tokens). This paper describes the changes made to the system to improve its perfor- mance by using the concept of layers and also by reducing the number of variables contained in the rules. It also describes the changes in rule syntax, such as the addition of new operators and contexts, which makes the rules more expressive. Resumo. Actualmente, a cadeia de PLN do L 2 F tem um modulo que ´ substancialmente mais lento que os outros, o RuDriCo. O RuDriCo ´ um desambiguador morfologico baseado em regras que tambem permite alterar a segmentacao de texto. Este trabalho descreve os melhoramen- tos realizados, nomeadamente a introducao de novos operadores, a in- troducao do conceito de camada e a reducao do numero de variaveis usadas na especificacao das regras.