Acceso usuarios
Vera Lucía Strube de Vera, Paulo Ricardo Carneiro Abrahao, Iván Santa María Filho
Most Natural Language Processing (NLP) systems traditionally use a sequential architecture that represents the classical linguistic levels (lexical-morphological, syntactical, semantic, etc.), but they claim for huge and monolithic modules, that make the backtracking among such modules difficult. By using a distributed architecture, we take advantages such as the parallelism in the procedural levels of the system, the distribution of the memory resources like knowledge and methods, and the interaction among the modules, which can change information by message passing. This article puts emphasis on a distributed proposal for the dictionary structure, in a NLP distributed environment. This proposal is oriented through a distributed perception of language processing, in a multi-agent system. Knowledge associated to morphology, syntax and semantic is embedded in specific agents, which are members of a NLP agents society. We present the dictionary models used for each of those agents, and we show how they are integrated in a system which processes Portuguese sentences.
