Validating Data from Semantic Web Providers

Abstract

As the Linked Open Data and the number of semantic web data providers hugely increase, so does the critical importance of the following question: how to get usable results, in particular for data mining and data analysis tasks? We propose a query framework equiped with integrity constraints that the user wants to be verified on the results coming from semantic web data providers. We precise the syntax and semantics of those user quality constraints. We give algorithms for their dynamic verification during the query computation, we evaluate their performance with experimental results, and discuss related works.

Role-Based Access Control on Graph Databases

“…However, it does not consider different forms of communication, which is the main focus of [18]. In doing so, we extend some aspects of each of our previous work [6,15,18], while relaxing others, in order to propose an efficient access control mechanism based on query rewriting.…”

“…The query rewriting process of our current proposal extends [6,15] by considering positive constraints with more than one atom in their bodies, but it does not consider special cases of unification treated in [15]. Indeed, [15] proposes a detailed algorithm to deal with instantiated constraints (e.g.…”

“…Both cases impose different forms of atom matching which we do not consider in our tests. Negative constraints treated in [6,15] are out of the scope of this paper as well. Notice that [6,15] only deal with conjunctive queries.…”

“…Negative constraints treated in [6,15] are out of the scope of this paper as well. Notice that [6,15] only deal with conjunctive queries. In the current paper, conjunctive queries (the only ones considered in the rewriting process) are the basis of groupby queries for which we propose an implementation method in a distributed data environment.…”

“…Query Rewriting Process. Constraints are used for query rewriting, following the proposal in [6], but, here, a constraint may have more than one atom in its body. More precisely, the query rewriting process is based on the chase procedure, in which constraints in C are triggered by atoms in body(q 1 ).…”