Responsible Data Science: impartiality, accuracy, confidentiality and transparency of data

• Autor(a) principal: Andrade, Morgana Carneiro
• Data de Publicação: 2020
• Outros Autores: Gonçalez, Paula Regina Ventura Amorim, Berti Junior, Decio Wey, Baptista, Ana Alice, Coneglian, Caio Saraiva
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Informação & Informação
• Texto Completo: https://ojs.uel.br/revistas/uel/index.php/informacao/article/view/38467
• topic: Data science; Ethic; Big data; Responsible data science; Ciencia de Datos; Ética; Big Data; Ciencia Responsable de Datos; Ciência dos dados; Ética; Big data; Ciência Responsável dos Dados

Introduction: In the Big Data context, as an urgent need arises the application of individual and corporate rights and regulatory standards that safeguard privacy, impartiality, accuracy and transparency. In this scenario, Responsible Data Science emerges as an initiative based on the FACT guidelines, which correspond to the adoption of four principles: impartiality, accuracy, confidentiality and transparency. Objective: To address alternatives that can ensure the application of the FACT guidelines. Methodology: An exploratory and descriptive research with a qualitative approach was developed. Searches were performed on the Web of Science, Scopus, and Scholar Google search engines using Responsible Data Science, Fairness, Accuracy, Confidentiality, Transparency Data Science, FACT, and FAT related to Data Science. Results: Responsible Data Science emerges as an initiative based on the FACT guidelines, which correspond to the adoption of the principles: impartiality, accuracy, confidentiality and transparency. In implementing these guidelines, consideration should be given to the use of techniques and approaches being developed by Green Data Science. Conclusions: It is concluded that Green Data Science and the FACT guidelines contribute significantly to safeguarding individual rights and that no measures need to be taken to prevent access and reuse of data. Challenges for implementing the FACT guidelines require studies, sine qua non conditions for tools for data analysis and dissemination to be developed at the design stage of methodologies.