Citation Context

...ss queries against federations of SPARQL endpoints with replicated fragments. 1 Scene setting The Linked Data Cloud includes more than one thousand datasets.1 However, a larger number of sources is available in the Web [14], and because of their heterogeneity their usage by Semantic Web technologies and in combination with Linked Data is limited. For the Web context, Local-as-View (LAV) data integration paradigm is well suited [1], nevertheless the traditional techniques used to produce query answers, query rewritings, may be too expensive in the context of SPARQL queries and numerous sources [21]. Many of the Linked Data datasets provide SPARQL endpoints.2 These endpoints allow users to explore datasets, and even use federated SPARQL query engines to answer queries using linked data across several datasets, i.e., federated queries. Unfortunately as the number and complexity of queries posed to the endpoints increases, their availability decreases [27]. An alternative to improve data availability is to give data consumers a more active role, and use their resources to replicate data, and consequently increase the data availability [18]. As Linked Data consumers are autonomous participa...