Análisis de sentimientos en base a las opiniones que realizan los usuarios en Twitter mediante aprendizaje automático

Abstract

Abstract: This degree work focuses on artificial intelligence, specifically on analysis of sentiments applied to opinion mining on Twitter. The developed web application allows you to analyze and classify tweets according to their polarity (positive or negative), with the aim of to know the opinion of users about the services and careers offered by the UTPL. This will help the institution to identify nonconformities and develop strategies for more effective communication. The platform allows the user to enter a hashtag (#) or a Twitter username (@) to extract and analyze tweets, using machine learning techniques. for your development, the SCRUM methodology was used, along with the Tweepy library to access the tweets and TextBlob to classify polarity using Naive Bayes models. Finally, Machine learning evaluation metrics were applied to measure performance in polarity classification.