Fake News Websites

Diogo Miguel Ferreira Pinheiro

Key information

Authors:

Diogo Miguel Ferreira Pinheiro (Diogo Miguel Ferreira Pinheiro)

Supervisors:

Maria Luísa Torres Ribeiro Marques da Silva Coheur (Maria Luísa Torres Ribeiro Marques da Silva Coheur); João Paulo Baptista de Carvalho (João Paulo Baptista de Carvalho)

Published in

11/24/2021

Abstract

Fake News have been around us for a long time. With the sudden rise of social media and influx ofinformation from various sources most people can not differentiate a truth from a lie. In this thesis,we approach this problem, focusing in Portuguese news websites. We try to find key elements thatdistinguish real news websites from fake websites. In this thesis, we proposed a system capable ofdistinguish trustworthy websites from websites that spread fake, non verified news. The system startsby receiving an url. Then an http request is made and the system tries to find the sub-page with theweb news site information. That page is processed and information is extracted. The system alsofinds other features, like the country localization and the provider. After gathering the features, theevaluator calculates the website score by giving a weight to each feature, marking the website as fakeor trustworthy.We created a data set with fake and trustworthy websites to test our system.We evaluated our system with different machine learning algorithms and concluded that the mostimportant features are the country that they are hosted and the security protocol.