A difusão de notícias sobre o encontro do Fórum Econômico Mundial em 2022: uma abordagem exploratória com Machine Learning e Deep learning

Abstract

The World Economic Forum is an international organization focused on international public-private cooperation. Since 1970, the institution has undergone several changes that have led it to have international recognition and prestige among international elites. According to Khanna, the Forum is a norm-setter, facilitator, and convener in a stakeholder-based structure. Through its annual meeting in Davos (Switzerland) the Forum receives great worldwide attention as it brings together leaders of public organizations of international organizations, private corporate representatives, activists, youth leaders, scientists, and the media. Consequently, a great deal of information is produced and made available on the Internet. Based on this data, the following research has the overall goal of exploring global news dissemination in terms of topics and prevailing sentiments. All methodology used is based on Machine Learning and Deep Learning, focusing on models from the LDA family, for topic modeling, and BERT for emotion analysis. The results suggest that the Corpus can be divided into 9 topics, and that primary, secondary and third interest can vary across countries. Furthermore, it was observed that the predominant feeling was "Joy", followed by "Anger". In the end, it could also be found that more than half of the Corpus from each country can be explained in up to three predominant topics.