Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/96770
Title: Análisis de sentimientos aplicado a la opinión política en Twitter: un sistema de clasificación en tiempo real
Author: Llorente Ayuso, Pablo
Tutor: Isern, David  
Others: Ventura, Carles  
Abstract: In this work we have developed a web application that shows, in real time, the polarized opinion of users on Twitter in Spanish about four political issues: feminism, LGBT rights, migrations and public services. The application is part of Sentiment Analysis discipline and opinions are obtained using a set of supervised machine learning algorithms. The preparation of the training data has been carried out by implementing symbolic natural language processing techniques through the use of functions with strings and regular expressions on the text of the tweets. For the selection of attributes, the inverse document frequency has been chosen over the words in the text. The final classification of an opinion results in the selection of the majority class of the set of classifiers. The results of the evaluation show an effectiveness between 77% and 94% depending on the selected metric and the political issue studied. The algorithm with the best results is the Bayesian classifier, although its efficiency converges with the classifier by majority as the distribution of the classes of the training set becomes more uniform. The architecture of the web application has three levels: presentation, business logic and integration, and it follows a model-view-controller pattern in a WAMP environment (Windows, Apache, MariaDB and PHP) in which the results of the classification are offered in JSON format using an API.
Keywords: sentiment analysis
Twitter
political opinion
Document type: info:eu-repo/semantics/bachelorThesis
Issue Date: 4-Jun-2019
Publication license: http://creativecommons.org/licenses/by-nc-nd/3.0/es/  
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
pllorenteayusoTFG0619memoria.pdfMemoria del TFG1 MBAdobe PDFThumbnail
View/Open