Prototipo para Procesamiento de Lenguaje Natural de documentos en español de diversas fuentes

Abstract

Abstract: Natural Language Processing investigates communication between machines and people in their context and language.The technological progress of the Cloud platforms to carry out PLN tasks has allowed the development of different solutions, although limitations are seen in terms of the analysis of languages other than English and text processing of different formats,which prevents the Cloud platforms from being able to comply with the expectations of analysis of documents in Spanish.The purpose of this paper is to establish the facilities provided by Cloud platforms, in particular for documents of various formats inSpanish. Some of the cloud platforms have been investigated, and among them the Google Cloud and Amazon Comprehend architectures have been chosen, to perform entity recognition and topic modeling tests, using format documents. In addition, other tests were carried out with improved architectures, including Python and NLTK libraries, to help with the preprocessing of the corpus and, in the case of Google, to display the results.