DETECTION OF FAKE NEWS IN THE SPANISH LANGUAGE USING MACHINE LEARNING TECHNIQUES

Nowadays, fake news has become a huge problem that causes damage around the world, especially in the social, political, and economic spheres. Due to the large amount of news generated every day, it is difficult to verify manually all the information to determine if a news item is real or fake. As a result, expert-based manual fact-checking, such as editors and journalists, need new tools that can perform the verification process efficiently. On the other hand, there are many studies focused on the detection of fake news in the English language, however, in the Spanish language, there are only a few researches that address this issue. For that reason, this proposed research explores different machine learning techniques to detect fake news in the Spanish language considering three feature extraction techniques: TF, TF-IDF, and Count Vectorizer; and five machine learning techniques: Logistic Regression, Stochastic Gradient Descent, Gradient Boosting, Random Forest and Support Vector Machine, were investigated and compared between them in order to achieve the classification task. Finally, the experimental results show the best performance with an accuracy rate of 87.18% using Random Forest as a classifier and TF as a feature extractor.