Uma assistente virtual para apoiar usuários finais nas atividades de Engenharia de Software utilizando o App Inventor

• Ano de defesa: 2022
• Autor(a) principal: Paiva, Francisco Gilberto Lima
• Orientador(a): Monteiro, Ingrid Teixeira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Desenvolvimento de usuário final; Assistente virtual inteligente; Aplicativos móveis; Engenharia de software
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/70422

One of the focuses of the Human-Computer Interaction area in recent years has been research focused on the field of End-User Development (EUD). Several works have been carried out to try to support users who are not professional developers to create their own systems. Consequently, as in systems developed by professionals, artifacts created by end users also need to present a certain level of quality. In order to contribute to this, research in End-User Software Engineering (EUSE) proposes to adapt Software Engineering (ES) traditional practices to end user development, so that their priorities are not changed. However, introducing awareness of best practices to the end-user is not so simple, due to their level of knowledge or their own intention to use the necessary effort for this. Therefore, this work presents the path towards the development of a virtual assistant with a voice interface integrated to App Inventor, a platform for the development of applications by end users through visual programming with blocks, aiming to help end users to practice EUSE activities in a EUD development environment. Studies were carried out with end users using App Inventor as well as an inspection of the screens of this platform. The resulting knowledge served as a basis for the construction of a virtual assistant that was later evaluated with new evaluations involving end users. The results allowed us to identify how higher education activities can be addressed by end users and how this process can be supported by a tool. Finally, we discuss how users considered the help of this virtual assistant to be positive and what points they need to improve from their perspectives. We also address how the App Inventor platform can support requirements, design, reuse, testing and debug activities, in addition to demonstrating some gaps that can be improved in this EUSE support process.