Sistema de recomendação de processos para o desenvolvimento de jogos digitais

• Ano de defesa: 2017
• Autor(a) principal: Politowski, Cristiano
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Ciência da Computação; UFSM; Programa de Pós-Graduação em Informática; Centro de Tecnologia
• Palavras-chave em Português: Processo de desenvolvimento de software; Desenvolvimento de jogos; Sistema de recomendação; Software development process; Game development; Recommendation system; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.ufsm.br/handle/1/13859

The digital gaming market is a billion dollar industry that faces issues in the way games are developed. One method to easy these issues is to use developer aid tools such as Recommendation Systems. These applications have a branch for software engineering, assisting the developer by generating recommended tasks. This work, therefore, aims to mitigate issues in digital game projects, developing a processes recommendation system, based on learning through past experiences in similar game projects, found in the form of postmortems. Our work brings three main contributions. The first one is a database containing experiences extracted from postmortems in the shape of development processes. The second is the definition of a context for digital gaming projects that characterizes and categorizes gaming projects. The third is a recommendation system for digital game projects that uses a list of similar projects and the development team profile to generate a new process with tips and guidance on tasks to be performed. Recommended processes assist developers in designing the game development process, listing activities and practices of projects with similar context, in the form of a software process. The extracted processes were validated with the game developers. The elements of the processes, generated by the recommendation system, were validated using precision and recall, and also together with the developers / authors of the analyzed game projects.