Título:	Do abstrato ao concreto : uma ferramenta para implementação de requisitos éticos em inteligência artificial através de histórias de usuário
Autor(es):	Borba, João Gabriel Rossi de
Orientador(es):	Canedo, Edna Dias
Assunto:	Ética Inteligência artificial - ética Engenharia de requisitos Requisitos éticos Aprendizagem de máquina Desenvolvimento de software
Data de publicação:	17-Mar-2025
Data de defesa:	28-Jan-2025
Referência:	BORBA, João Gabriel Rossi de. Do Abstrato ao Concreto: Uma Ferramenta para Implementação de Requisitos Éticos em Inteligência Artificial através de Histórias de Usuário. 2025. 129 f. Dissertação (Mestrado em Informática) — Universidade de Brasília, Brasília, 2025.
Abstract:	Context: In recent years, the field of artificial intelligence has undergone a remarkable process of expansion, both at the academic and industrial levels. This growth can be observed in various ways, including the development of more complex technologies, increased investment, greater media attention and the expansion of its areas of application. However, this advancement has also given rise to a number of ethical issues that are becoming a matter of growing social concern, including system bias and the potential for harmful applications. Objective: This work aims to develop a tool that operationalises the translation of high-level ethical requirements into ethical user stories. Methods: In order to achieve this objective, this study was based on the Design Science Research methodology, which has yielded four results. Initially, in the problem awareness phase, a systematic literature review was updated. Subsequently, in the suggestion phase, the proposed tool was formulated as a system that transforms ethical requirements into ethical user stories. Following the second phase, the development stage, the first version of the tool was implemented. Finally, the tool was subjected to evaluation in order to generate empirically based evidence of whether it is fit for purpose. Results: A total of 38 primary studies were identified. The majority of these studies (63%) put forward a practical solution to facilitate the application of ethics in AI. Nevertheless, a discrepancy persists between theoretical and practical applications of ethics in AI. Additionally, a list of 26 frequently discussed ethical principles in the literature was compiled. Subsequently, the Requirements to US tool was proposed, developed, and validated. This tool employs ethical user stories, AI ethics principles, and Large Language Models to facilitate the integration of AI ethics during the requirements engineering phase. The validation process yielded positive outcomes, confirming the effectiveness of the tool. Conclusion: The outcomes of the model’s training process demonstrate its potential as a valuable tool for the early integration of AI ethics into software development. The results indicate that the tool supports the integration of theoretical and practical aspects, offering a promising solution to an existing gap in the current range of practical applications.
Unidade Acadêmica:	Instituto de Ciências Exatas (IE) Departamento de Ciência da Computação (IE CIC)
Informações adicionais:	Tese (doutorado)—Universidade de Brasília, Instituto de Ciências Exatas, Departamento de Ciência da Computação, 2025.
Programa de pós-graduação:	Programa de Pós-Graduação em Informática
Licença:	A concessão da licença deste item refere-se ao termo de autorização impresso assinado pelo autor com as seguintes condições: Na qualidade de titular dos direitos de autor da publicação, autorizo a Universidade de Brasília e o IBICT a disponibilizar por meio dos sites www.unb.br, www.ibict.br, www.ndltd.org sem ressarcimento dos direitos autorais, de acordo com a Lei nº 9610/98, o texto integral da obra supracitada, conforme permissões assinaladas, para fins de leitura, impressão e/ou download, a título de divulgação da produção científica brasileira, a partir desta data.
Aparece nas coleções:	Teses, dissertações e produtos pós-doutorado

Os itens no repositório estão protegidos por copyright, com todos os direitos reservados, salvo quando é indicado o contrário.