Aplicación del análisis de sentimientos a la evaluación de datos generados en medios sociales

Abstract

This document describes the process of research and development conducted in the discipline of sentiment analysis. The main goal of this research was to evaluate the application of sentiment analysis technologies to user-generated content from social media and to outline proposals for the use of the results of these technologies to organizations and users. The extent of reliability of online tools sentiment analysis working with Twitter as a source of corpus was studied; a heuristic approach that simplifies the sentiment analysis process of Twitter messages focusing on the opinions directly related to targets of opinion rather than determining sentiment globally and generates additional information which might be useful for electronic word of mouth was presented; Finally a proposal for quantitative prediction of hotels’ ratings from reviews by users of their services was developed and evaluated. The results of this research show that sentiment analysis is a discipline that in its current state can be helpful in making decisions for companies and individuals but it is still improvable to take advantage of the massive amount of opinions in text issued by users of social media