Using Natural Language Processing in the design of a multilingual architecture for representation of Language

Abstract

Researches on models and architectures of multilingual information retrieval have increased in recent years mainly due to the need to exchange knowledge and information between different cultures. In regards to this need, this article proposes multilingual architecture techniques using Natural Language Processing (NLP). In addition, the article presents the use of morphological, syntactic and semantic analyses as the fundamental contribution for this model; also, an alternative set out the problem of multilingualism, which is basically the use of a technique for universal representation of a language by labeling the lexical components with grammatical features, and the reuse of language resources such as ILI of EuroWordNet. This proposal is the result of a study and analysis of current trends in research and existing technologies related to this knowledge area. © 2010 IEEE.