Abordagem inteligente para alocação de tarefas em projetos de software

• Ano de defesa: 2022
• Autor(a) principal: Fensterseifer Filho, Evandro Luis da Rosa
• 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 Ciência da Computação; Centro de Tecnologia
• Palavras-chave em Português: Sistemas de recomendação; Alocação de tarefas; Gerenciamento de projetos; Recommender systems; Task allocation; Project management; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.ufsm.br/handle/1/25786

The team's technical knowledge and personality influence software projects, and can reduce or increase the quality of the generated products and the speed of development. For successful task allocation, it is needed to consider the preferences and profile of each developer, thus maximizing their productivity. In projects with many developers, task allocation can be a challenging task and can be aided by recommendation tools. In this work, we propose an intelligent approach for allocating software development tasks appropriate to the developer's profile. Based on the literature, profiles of developers needed in a software team were defined, considering skills and technical profiles. Aiming to evaluate the profile of the developer and associate appropriate tasks with it, this work uses a questionnaire with questions that aim to identify the profile of the developer. From the answers, a recommender system allocates tasks to developers, employing text processing techniques, MultinomialNB, and Random Forest. Developers evaluate allocations, and the system uses them to improve the recommender system. The allocation of tasks, according to the profile of each developer, seeks to improve the performance of the project and the quality of the products developed, as well as to reduce the development effort. The validations showed that the developed approach makes consistent and coherent task recommendations to developers, according to their profile.