Multidimensional Enrichment of Spatial RDF Data for SOLAP -- Full Version

TL;DR

This paper introduces RDF2SOLAP, an enrichment model that automatically annotates spatial multidimensional data with QB4SOLAP, enabling SOLAP queries over existing RDF data on the Semantic Web.

Contribution

It proposes and evaluates enrichment algorithms that bridge the gap between QB/QB4OLAP and QB4SOLAP for spatial data analysis.

Findings

Enrichment algorithms successfully annotate spatial data for SOLAP.

Application on real-world government data demonstrates practical utility.

Enhanced query capabilities over spatial RDF datasets.

Abstract

Large volumes of spatial data and multidimensional data are being published on the Semantic Web, which has led to new opportunities for advanced analysis, such as Spatial Online Analytical Processing (SOLAP). The RDF Data Cube (QB) and QB4OLAP vocabularies have been widely used for annotating and publishing statistical and multidimensional RDF data. Although such statistical data sets might have spatial information, such as coordinates, the lack of spatial semantics and spatial multidimensional concepts in QB4OLAP and QB prevents users from employing SOLAP queries over spatial data using SPARQL. The QB4SOLAP vocabulary, on the other hand, fully supports annotating spatial and multidimensional data on the Semantic Web and enables users to query endpoints with SOLAP operators in SPARQL. To bridge the gap between QB/QB4OLAP and QB4SOLAP, we propose an RDF2SOLAP enrichment model that automatically annotates spatial multidimensional concepts with QB4SOLAP and in doing so enables SOLAP on existing QB and QB4OLAP data on the Semantic Web. Furthermore, we present and evaluate a wide range of enrichment algorithms and apply them on a non-trivial real-world use case involving governmental open data with complex geometry types.