Uso de análise de sentimentos no estudo da reação de usuário ao conteúdo de vídeos no Youtube

With the emergence of new technologies, the use of digital media has been growing, making this type of media important for companies and the market. The relationship and user engagement with content have been the object of interest in several studies, in particular with a focus on the ability to ass...

Descripción completa

Detalles Bibliográficos
Autor: Carvalho, Luiz Antônio de
Tipo de recurso: tesis de maestría
Estado:Versión publicada
Fecha de publicación:2022
País:Brasil
Institución:Escola Superior de Propaganda e Marketing (ESPM)
Repositorio:Biblioteca Digital de Teses e Dissertações da ESPM
Idioma:portugués
OAI Identifier:oai:tede2.espm.br:tede/656
Acceso en línea:http://tede2.espm.br/handle/tede/656
Access Level:acceso abierto
Palabra clave:youtube; vídeos online; análise de sentimento; mineração de dados; biblioteca de sentimentos
youtube; online videos; sentiment analysis, data mining, sentiment library
CIENCIAS SOCIAIS APLICADAS::ADMINISTRACAO
Descripción
Sumario:With the emergence of new technologies, the use of digital media has been growing, making this type of media important for companies and the market. The relationship and user engagement with content have been the object of interest in several studies, in particular with a focus on the ability to assess user response to content. The objective of this final paper is to study the relationship between the polarity of feelings contained in the comments of videos on Youtube and the particular characteristics of the videos produced, such as the content and aspects of the message and the interlocution. For this, a process was developed to extract and evaluate the positive and negative feelings expressed in comments on Youtube videos of a Brazilian online course company, which operates in the field of public tenders. Through the analyzed data, it was possible, for the videos with the highest viewing of the channel, to identify the relationships between the feelings exposed in the comments and the content of the videos. It is hoped that this final paper can contribute to the generation of inputs to improve the quality of the production of new videos and the provision of services to users of Youtube channels.