Analyzing Scopus articles on post-truth and fake news with an unsupervised learning algorithm

Bohorquez-Lopez V.W.; Gomez-Burns A.E. · 28th Americas Conference on Information Systems, AMCIS 2022 (2022)

Resumen

The terms post-truth and fake news have been studied by numerous authors, using various approaches and methodologies. In this context, we seek to consolidate everything published in Scopus on the subject, proposing the following objectives: identify the most frequent topics per year, related to post-truth and fake news; and classify the topics obtained, taking into account their evolution over the years. The methodology used consisted of applying topic modeling, which uses an unsupervised learning algorithm, extracting 9 important topics, which were later classified into two blocks: proliferation of fake news in political, public and journalistic contexts; and, misinformation and disinformation, and how we can combat it. The contributions of the study are: the methodology that can be replicated in other contexts; and the topics and classifications obtained that can serve as a starting point for future research on post-truth and fake news.

Instituciones: Universidad Católica San Pablo