| Resumen |
There are a number of works that deal with the classification of feelings evoked by various types of documents, from literaty texts to social networks like Twitter. Various classifications can be performed based on the sentiments considered by the author. The goal of this article is to classify a collection of tweets in different contexts in which the word 'happy' or 'happiness' can be used, for example advertising, congratulations or as a simple sarcasm. This will be done using supervised learning systems. Several natural language processing methods such as tokenization, functional word identification, and n-grams will be employed. © 2022 Universidade do Minho. All rights reserved. |
