| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.63 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
No contexto de uma sociedade em constante evolução tecnológica, o reconhecimento da privacidade como um requisito essencial tem se intensificado de maneira significativa. À medida que a tecnologia avança e os dispositivos se tornam cada vez mais interconectados, a proteção dos dados pessoais e a segurança da informação têm assumido uma importância fundamental. A importância da proteção de dados torna-se ainda mais crucial no contexto das informações e operações militares, destacando a segurança nacional, as missões críticas e a proteção dos militares, militarizados e civis da organização.
Este trabalho tem como objetivo o estudo de métodos de Controlo de Divulgação Estatística (CDE) aplicados a bases de dados do foro naval, isto é, uma base de dados que engloba dados sobre embarcações de pesca e fiscalizações realizadas às mesmas por navios da Marinha Portuguesa. Para tal, foram utilizados dados reais, que foram transformados para garantir a sua proteção. Posteriormente, foi analisada a qualidade dos dados protegidos comparando-os com os dados originais. Nestes dados foram usadas as técnicas de adição de ruído, arredondamento, recodificação global, supressão local, imputação parcial e deslocamento geográfico.
Com base nos resultados obtidos, neste trabalho foi possível demonstrar a eficácia das técnicas de proteção escolhidas através da comparação de histogramas de variáveis originais e protegidas, comparação de médias e variâncias, e ainda, através de matrizes de correlação de Pearson, Kendall e Point-Biserial.
A qualidade demonstrada dos dados neste trabalho contribuiu para o avanço do conhecimento sobre o CDE e das estratégias de proteção de dados na Marinha Portuguesa, estimulando investigações futuras no âmbito deste tema.
In a society constantly evolving technologically, the recognition of privacy as an essential requirement has significantly intensified. As technology advances and devices become increasingly interconnected, the protection of personal data and information security have taken on fundamental importance. The importance of data protection becomes even more crucial in the context of military information and operations, highlighting national security, critical missions and the protection of the organization’s military, militarized and civilian personnel. The objective of this work is the study of Statistical Disclosure Control (SDC) methods applied to naval databases, that is, a database that includes data on fishing vessels and inspections carried out by Portuguese Navy ships. For this, real data was used, which was transformed to guarantee its protection. Subsequently, the quality of the protected data was analyzed by comparing it with the original data. In these data, noise addition, rounding, global recoding, local suppression, partial imputation and geographic displacement techniques were used. Based on the results obtained, in this work it was possible to demonstrate the effectiveness of the protection techniques chosen through the comparison of histograms of original and protected variables, comparison of means and variances, and also, through correlation matrices of Pearson, Kendall and Point-Biserial. The demonstrated quality of the data in this work contributed to the advan cement of knowledge about the SDC and data protection strategies in the Portuguese Navy, stimulating future investigations within the scope of this topic.
In a society constantly evolving technologically, the recognition of privacy as an essential requirement has significantly intensified. As technology advances and devices become increasingly interconnected, the protection of personal data and information security have taken on fundamental importance. The importance of data protection becomes even more crucial in the context of military information and operations, highlighting national security, critical missions and the protection of the organization’s military, militarized and civilian personnel. The objective of this work is the study of Statistical Disclosure Control (SDC) methods applied to naval databases, that is, a database that includes data on fishing vessels and inspections carried out by Portuguese Navy ships. For this, real data was used, which was transformed to guarantee its protection. Subsequently, the quality of the protected data was analyzed by comparing it with the original data. In these data, noise addition, rounding, global recoding, local suppression, partial imputation and geographic displacement techniques were used. Based on the results obtained, in this work it was possible to demonstrate the effectiveness of the protection techniques chosen through the comparison of histograms of original and protected variables, comparison of means and variances, and also, through correlation matrices of Pearson, Kendall and Point-Biserial. The demonstrated quality of the data in this work contributed to the advan cement of knowledge about the SDC and data protection strategies in the Portuguese Navy, stimulating future investigations within the scope of this topic.
Description
Keywords
Controlo de Divulgação Estatística Privacidade Proteção de Dados Divulgação Análise de Padrões