| Abstract |
A cornerstone issue in the realisation of the Semantic Web (SW) vision is the achievement of semantic interoperability among legacy data sources spread worldwide. In order to capture information semantics in a machine processable way, various ontology-based formalisms have been recently proposed (e.g., RDF/S). However, the vast majority of existing legacy data is not yet in RDF/S or any other SW language. As a matter of fact, most of the data is physically stored in relational database (RDB) systems and published on the Web as XML. SW applications, however, require to view data as virtual RDF, valid instance of a domain or application specific RDF/S schema, and to be able to manipulate them with high-level query languages, such as RQL. Therefore, we propose a middleware system that allows querying RDB data using RQL. So, our work focuses on specifying a first-order encoding for RDF, namely SWLF, along with constraints preserving RDF/S semantics that will allow specifying RDB to RDF mappings, composing RQL queries with these mappings - thus, producing RDB queries (a.k.a query reformulation) - and performing query optimisations (minimisations). Special emphasis is given on RQL query containment and optimisation. Although there is limited background work on RQL query optimisation (and containment), it is very important since it allows saving time and space, which are two very valuable resources when managing queries, especially over distributed systems. The optimisation removes RQL query redundancy (by taking advantage of the RDF/S constraints) and redundancy of the reformulated query (by exploiting constraints of the underlying RDB and other).
|
Βελτιστοποιώντας και Αναδιατυπώνοντας RQL Επερωτήσεις στο Σημασιολογικό Ιστό
