Privacy Everywhere: mecanismo para tomada de decisões e garantia da privacidade em ambientes IoT

• Ano de defesa: 2019
• Autor(a) principal: Andrade, Leandro Prado de
• Orientador(a): Zorzo, Sérgio Donizetti
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Português: Internet das coisas; Privacidade em IoT; Preferências de privacidade; Tomada de decisão
• Palavras-chave em Inglês: Internet of things; IoT Privacy; Privacy preferences; Decision making
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/12149

Through technological advances, society has moved toward the "always connected" paradigm. Growing varieties of devices are becoming able to connect to a network, capture information around them, and send and receive information. On the other hand, increasing in number of Internet of Things (IoT) devices, the ability to integrate IoT devices with IoT cloud services, and the diversity of forms of interaction can make it tiring to ask the user to make a privacy decision whenever it contacts a new IoT device. Due to sensitivity of many of information captured by IoT devices and because IoT cloud services do not focus on preserving user privacy, it is necessary that data captured by IoT devices be processed before being sent to cloud services. IoT environments can have a large number of data collection devices that often go unnoticed by users. Informing the user about the presence of each one of these devices and asking the user to make privacy decisions about data collection from these devices may be impracticable. This work presents the Privacy Everywhere mechanism for decision making and user privacy assurance in IoT environments. The proposed mechanism addresses user privacy issues and makes use of trained neural networks with data collected from a pre-established community. This minimizes user consultation at the time of data collection by the device that is part of an IoT environment. In validation of the mechanism, accuracy check of the Allow/Deny and Privacy Action neural networks that make up the mechanism presented an accuracy of 88.02% and 86.67%, respectively. Privacy Everywhere mechanism is able to help users preserve the privacy of their data in IoT environments.