Estudo de caso do algoritmo de recomendação do YouTube com base em análise de redes sociais

Abstract

With the increase in users using social networks, made possible by the popularization of technology and the Internet, it was necessary to study this area in order to understand phenomena such as the circulation of information, relationships, and even an increase in the dissemination of false information, since the dissemination of information or even products through these networks has great market value. The area that studies such phenomena is called Social Network Analysis, and aims to identify influential users in a network or community through analysis focused on relationships between entities. Such an analysis is done through concepts called actors, relationships and networks of connections and the transformation of such concepts into a graph, which will be explored through several algorithms. The objective of this work is to collect data from a social network (Youtube) and transform them into a graph, in order to expose and analyze the influence on social networks through social network analysis algorithms. As a result, a data collector for youtube was developed, which makes it possible to transform them into a graph, finally having their data analyzed through correlations, not identifying an association between video data (such as likes) and the result of the algorithms applied.