Adaptación de una herramienta de procesamiento de lenguaje natural para el etiquetado de sentimientos y el análisis de lenguaje en español. (Trabajo de Titulación de Ingeniero en Sistemas Informáticos y Computación ). UTPL, Loja

Abstract

The objective of this work was to adapt degree of tool Natural Language Processing and Labelling of Feelings in Spanish, based on the systematic and linguistic analysis of text. At present there are several programs that help with natural language processing (NLP), however in this work OpenNLP used due to the advantages of as: Consume fewer resources and performs processing in less time. OpenNLP a collection of projects released under open source license, developed in Java, which provides the following tools: tokenizer, detection prayers, recognizer of the entity and part-of-speech tagging. To obtain the data, first he split the text into sentences. Each sentence was split into tokens to assign a grammatical label. This grammatical label was proposed by the EAGLES group for European languages, including Spanish. In addition, the token was assigned labeling emotions as boredom, anxiety, anxiety or worry, confusion, frustration and sympathy. Finally, many of the tools that are available in Spanish are limited, therefore it wanted to implement a tool with enhanced functionality to the benefit of new research