SPARQL with XQuery-based Filtering

TL;DR

This paper introduces XQueryFILTER, an extension to SPARQL that enables seamless querying of RDF and XML data, with optimized query processing demonstrated through real-world experiments.

Contribution

The paper proposes XQueryFILTER, a novel extension to SPARQL for integrating XQuery filtering, along with a database theory-based optimization for efficient combined RDF and XML data querying.

Findings

XQueryFILTER effectively enables integrated RDF and XML data querying.

The proposed optimization improves query processing efficiency.

Experiments confirm the approach's effectiveness in real-world scenarios.

Abstract

Linked Open Data (LOD) has been proliferated over various domains, however, there are still lots of open data in various format other than RDF, a standard data description framework in LOD. These open data can also be connected to entities in LOD when they are associated with URIs. Document-centric XML data are such open data that are connected with entities in LOD as supplemental documents for these entities, and to convert these XML data into RDF requires various techniques such as information extraction, ontology design and ontology mapping with human prior knowledge. To utilize document-centric XML data linked from entities in LOD, in this paper, a SPARQL-based seamless access method on RDF and XML data is proposed. In particular, an extension to SPARQL, XQueryFILTER, which enables XQuery as a filter in SPARQL is proposed. For efficient query processing of the combination of SPARQL and XQuery, a database theory-based query optimization is proposed. Real-world scenario-based experiments in this paper showcase that effectiveness of XQueryFILTER and efficiency of the optimization.