Syntactic privacy in dynamic publication
In the domain of privacy-preserving data publication, conventional approaches have primarily emphasized the creation of unique releases. However, these approaches often fail to address the challenge of republishing microdata after updates that involve insertions and deletions. This limitation has no...
| Autor: | |
|---|---|
| Tipo de documento: | dissertação |
| Data de publicação: | 2024 |
| País: | España |
| Recursos: | Universitat Politècnica de Catalunya (UPC) |
| Repositório: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglês |
| OAI Identifier: | oai:upcommons.upc.edu:2117/422145 |
| Acesso em linha: | https://hdl.handle.net/2117/422145 |
| Access Level: | Acceso aberto |
| Palavra-chave: | Data protection privacy m-invariance Protecció de dades Àrees temàtiques de la UPC::Informàtica::Seguretat informàtica |
| Resumo: | In the domain of privacy-preserving data publication, conventional approaches have primarily emphasized the creation of unique releases. However, these approaches often fail to address the challenge of republishing microdata after updates that involve insertions and deletions. This limitation has notable implications, as it hinders publishers from providing continuously updated datasets to researchers. Fortunately, an innovative approach documented in the literature offers a promising solution to this limitation. At its core, this solution is grounded in the novel concept of m-invariance, a generalization principle. This thesis aims to address the lack of a publicly available code implementing the m-invariance principle, followed by a comprehensive analysis of the results obtained through its implementation. The algorithm was tested using a real database. The results indicate that the published data complies with the principle of m-invariance. In terms of code efficiency, the execution times are within acceptable limits. The analysis of the counterfeits reveals a higher rate of addition for larger dataset sizes or higher m values. However, as the update volume increases, the number of counterfeits added decreases. |
|---|