An approach for a security and privacy-aware cloud-based storage of data in the Semantic Web
Jens Köhler, Thomas Specht, Kiril Simov · 2016
This work proposes to include security and privacy into to context of Linked Open Data (LOD) and the Semantic Web. Here, a database partitioning approach to logically and physically separate data and distribute the partitions across several different cloud providers is applied. This data distribution is called fixed vertical partitioning and distribution approach (FVPD), as data schemes are vertically partitioned and distributed across various different cloud providers. Hence, this framework uses a security-by-distribution approach to storing sensitive data securely and privacy-ware in all possible cloud deployment models and thus, it provides a solution to improve the level of security and privacy. The proposed FVPD approach also becomes viable in the context of LOD and the Semantic Web. Furthermore, in relevant literature no approach in this context has been dealing with security and privacy so far. Therefore, this work transfers the FVPD approach from relational databases to Semantic Web Frameworks (Apache Jena, Sesame, etc.) that provide storage engines for RDF data and corresponding SPARQL endpoints. Above that, the query performance of such a federated environment is compared to a non-partitioned and non-distributed environment. For this, an adapted version of the TPC-W benchmark is used. Thus, this work shows how the query performance is affected when relational database data are exposed as SPARQL endpoints in RDF-based form and how the FVPD approach influences the query performance. Finally, it illustrates a mechanism to improve the level of security and privacy through a security-by-distribution approach.