Exposing vocabularies for soil as Linked Open Data

In a recent paper published in the journal Information Processing in Agriculture, entitled  "Exposing vocabularies for soil as Linked Open Data"  . Giovanni L’Abate , Caterina Caracciolo , Valeria Pesce , Guntram Geser , Vassilis Protonotarios , and  Edoardo A.C. Costantini review standards for describing soil data and reports on the work done with the EC funded agINFRA project to apply Linked Data technologies to existing standards and data in order to improve the interoperability of soil datasets.

The paper sets the tone by detailing current developments in data interoperability and  'semantic interoperability' - and further provides AGRIS and AGROVOC as the examples of the adoption of the linked data approach in the agricultural domain. Thereafter, the paper collocates the agINFRA work which focussed on identifying and recommending existing RDF vocabularies and publishing new ones. agINFRA aims to provide the tools and methodology to be used for the publication of the data managed by project partners as Linked Open Data (LOD). This is expected to significantly facilitate the interoperability between heterogeneous data sources, not previously linked.

The main results of the work described in the paper is twofold - establishing an RDF Vocabulary for soil concepts based on the UML INSPIRE model, and secondly, having a Knowledge Organization System for soil data published and mapped to existing relevant KOS, based on the analysis of the SISI database of the CREA of Italy.

The original SISI relational database was mapped to different micothesauri and their concepts. Mappings of each concepts were searched in the INSPIRE Registry, the NAL Thesaurus and the Linked Thesaurus Framework for Environment (LusTRE). When an exact match was found, the descriptor from INSPIRE or the Agricultural Thesaurus was adopted as an alternative English label of the concept. The resultant RDF Soil Vocabulary was based on the INSPIRE model and the SISS web application, and the publication of an RDF Soil Concepts KOS. The editing tool used to maintain the Soil Concept KOS was the VocBench.

This work provided a major contribution in that it proposed and applied a methodology to deal with the conversion of small sized KOS, used in local databases, into linked vocabularies.

For further information read the paper here