| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 660.58 KB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Os sistemas de recomendação são ferramentas de inteligência artificial de filtragem de informações que buscam melhorar a experiência individual online ao indicar itens personalizados de acordo com as preferências, comportamentos e interesses extraídos dos dados dos usuários. Apesar de potencialmente benéfica, a atual forma de funcionamento cria riscos aos direitos humanos. Amparado em análise de doutrina jurídica internacional, busca-se possíveis soluções e estratégias de mitigação dos efeitos maléficos causados pelos sistemas de recomendação, principalmente em termos de violações de privacidade, diminuição da autonomia, manipulação, enfraquecimento de processos democráticos e discriminação. Nesse ambiente complexo, além da regulação jurídica em termos de proteção de dados, uma abordagem multidisciplinar e multisetorial, com base em princípios éticos, medidas educacionais, soluções tecnológicas e estratégias mercadológicas, que colocam o ser humano no centro da questão, parece ser ideal para minimizar os riscos e potencializar os benefícios da utilização desses sistemas.
Recommender systems are artificial intelligence tools for filtering information that seek to improve the individual online experience by recommending personalized items, based on preferences, behaviours and interests extracted from user’s data. Although potentially beneficial, the current implementation of these systems creates risks to human rights. Supported by an analysis of international legal doctrine, this work intends to seek possible solutions and strategies to mitigate the harmful effects caused by recommendation systems, mainly in terms of privacy violations, decreased autonomy, manipulation, impairment of democratic processes and discrimination. In this complex environment, in addition to legal regulation in terms of data protection, a multidisciplinary and multisectoral approach, based on ethical principles, educational measures, technological solutions and marketing strategies, which place the human being at the centre of the issue, seems to be ideal for minimize risks and enhance the benefits of using these systems.
Recommender systems are artificial intelligence tools for filtering information that seek to improve the individual online experience by recommending personalized items, based on preferences, behaviours and interests extracted from user’s data. Although potentially beneficial, the current implementation of these systems creates risks to human rights. Supported by an analysis of international legal doctrine, this work intends to seek possible solutions and strategies to mitigate the harmful effects caused by recommendation systems, mainly in terms of privacy violations, decreased autonomy, manipulation, impairment of democratic processes and discrimination. In this complex environment, in addition to legal regulation in terms of data protection, a multidisciplinary and multisectoral approach, based on ethical principles, educational measures, technological solutions and marketing strategies, which place the human being at the centre of the issue, seems to be ideal for minimize risks and enhance the benefits of using these systems.
Description
Keywords
Inteligência artificial Sistemas de recomendação Perfilamento Decisões automatizadas Direitos humanos Proteção de dados Artificial intelligence Recommender systems Profiling Automated decision-making Human rights Data protection