Citation Context

... Romero 2012), or on queries expressed in less expressive query languages (for example without self-joins) (Fink and Olteanu 2011), or on queries for which the answers have been pre-computed (views) (=-=Halevy 2001-=-). On the other hand, there is a lot of work for query extension using semantic data descriptions, but mainly in the area of information retrieval focusing more on the representation of the query cont...

Citation Context

...he query execution times compared to evalStatic. We evaluated evalStatic and evalExt over the Lehigh University Benchmark (LUBM) (Guo, Pan, and Heflin 2005), the University Ontology Benchmark (UOBM) (=-=Ma et al. 2006-=-), and the Semintec ontology from the Oxford ontology library.1 We used a range of custom queries since the queries provided for LUBM and UOBM are only simple conjunctive instance queries. Additionall...

Citation Context

...gy to extend the query such that the additionally added parts allow for better query planning and an overall faster query evaluation. Compared to query evaluation with standard static query planning (=-=Kollia and Glimm 2013-=-) our empirical analysis with a prototypical implementation shows an improvement of up to four orders of magnitude in the query execution times due to the proposed optimizations. The idea of modifying...

Citation Context

...o work well in practice, Zhou et al. also propose optimizations such as extracting relevant fragments of the ontology for checking a potential answer and an adaptation of the summarization technique (=-=Dolby et al. 2009-=-; 2007). In this paper, we also follow the hybrid approach, but we significantly extend the form of allowed queries. Furthermore, we present a way of using the knowledge in the ontology to extend the ...

Citation Context

...irect Semantics entailment regime of SPARQL 1.1. Kollia et al. (Kollia and Glimm 2013) propose a method (referred to as evalStatic) for computing bounds for query templates using the HermiT reasoner (=-=Glimm et al. 2014-=-). In particular, the classified concept hierarchy and the premodel that HermiT constructs are used to compute lower and upper query template bounds as described in Example 3. Using these bounds a sop...