Flexible query processing for SPARQL

Abstract

Flexible querying techniques can enhance users' access to complex, heterogeneous datasets in settings such as Linked Data, where the user may not always know how a query should be formulated in order to retrieve the desired answers. This paper presents query processing algorithms for a fragment of SPARQL 1.1 incorporating regular path queries (property path queries), extended with query approximation and relaxation operators. Our flexible query processing approach is based on query rewriting and returns answers incrementally according to their "distance" from the exact form of the query. We formally show the soundness, completeness and termination properties of our query rewriting algorithm. We also present empirical results that show promising query processing performance for the extended language.

Foundations of Modern Query Languages for Graph Databases

“…Various such extensions are concerned with supporting additional meta-information for RDF data: two such proposals are SPARQL* [Hartig and Thompson 2014] and AnQL [Zimmermann et al 2012], which both describe general frameworks for reifying or annotating RDF data (respectively), providing analogous query features in SPARQL. Other general extensions of interest include SPARQL AR [Frosini et al 2017], which allows for performing query approximation and relaxation to also return "near answers"; SPARQLog [Bry et al 2009], which extends SPARQL with rules and more flexible forms of quantification, additionally defining fragments that maintain desirable complexity results; XSPARQL [Bischof et al 2012], which allows for federating queries over SPARQL, XML (through XQuery) and relational databases (through SQL) in a unified manner; as well as work by Lausen et al [2008] on using SPARQL (and a proposed extension thereof) to specify relational-like constraints over RDF graphs.…”

Answers Partitioning and Lazy Joins for Efficient Query Relaxation and Application to Similarity Search

“…The existing approaches for query relaxation consist in enumerating relaxed queries up to some edit distance, and to evaluate each relaxed query, from the more specific to the more general, in order to get new approximate answers [15,14,9]. The main problem is that the number of relaxed queries grows in a combinatorial way with the edit distance, and the size of the query.…”

“…Hurtado et al [15] optimize the evaluation of relaxed queries by directly computing their proper answer but the optimization only works for the relaxations that replace a triple pattern by a single other triple pattern. They also introduce new SPARQL clauses, RELAX and APPROX [9], to restrict relaxation to a small subset of the query. However, this requires from the user to anticipate where relaxation can be useful.…”

“…We note N orm(P ) the result of this normalization on pattern P . The node-to-variable relaxation then amounts to the removal of one query element, filter x n = n. For the sake of simplicity, we do not include here the relaxation of property paths [9]. Another kind of relaxation that we let to future work is the gradual relaxation of equality filters.…”

“…It consists in applying transformations to the user query in order to relax constraints, and make it more general so that it produces more answers. In previous work on SPARQL queries over RDF(S) graphs [15,5,6,14,9], typical relaxation steps consist in removing a query element or replacing a class by a superclass or a node by a variable. The major limitation of query relaxation is that the number of relaxed queries grows in a combinatorial way with the number of relaxation steps and the size of the query.…”