Presentation delivered at the Agricultural Data Interoperability Interest Group -- Research Data Alliance (RDA) 4th Plenary Meeting -- Amsterdam, September 2014
The agINFRA Linked Data layer by Valeria Pesce, Giovanni l'Abate, Luca Matteis and Koraljka Golub
1. The agINFRA Linked Data layer
Valeria Pesce
Global Forum on Agricultural Research (GFAR)
Giovanni L’Abate
Consiglio per la Ricerca e la sperimentazione in Agricoltura
Centro di ricerca per l’agrobiologia e la pedologia (CRA-ABP)
Luca Matteis
Koraljka Golub
Research Data Alliance 4th Plenary Meeting
22-24 September 2014, Amsterdam
Agricultural Data Interoperability Interest Group
agINFRA project
EC 7th framework program INFRA-2011-1.2.2 - Grant agr. no: 283770
2. agINFRA - Background
• agINFRA: FP7 project
EC 7th framework program INFRA-2011-1.2.2 - Grant agr. no: 283770
Objective: Promoting data sharing and development
of trust in agricultural sciences
• agINFRA Knowledge Fair co-located with the
Agricultural Data Interoperability Interest group
meeting at the RDA 4th Plenary Meeting
• Types of data covered by agINFRA:
bibliographic, educational, germplasm, soil
3. Interoperability 1
They are often both called vocabularies
Metadata elements
to describe individual “things” (entities, datums, series…)
Aka metadata sets, metadata element sets, vocabularies
Sets of values
for (some of) the metadata elements
Aka controlled vocabularies, authority data, value
vocabularies, Knowledge Organization Systems (KOSs)
4. Various flavors of vocabularies
Title
Author(s)
Abstract
Subject(s)
Publication date
Publication place
Type of document
other features…
Entity to be describedType?
Bibliographic
resource
for describing
bibliographic
resources
Metadata
vocabulary
Authority data
KOS
“Value vocabularies”
Data of type Person
Authority data
Data of type
Geographic location
“Description
vocabularies”
Controlled list
Concepts suitable for
organizing by Topic
Concepts suitable for
organizing by Type
for describing
people
Metadata
vocabularyfor describing
geographic places
Ontology
5. Names of
things
URIs of
things
Links to
other URIs
Metadata vocabularies RDFs / OWL
KOS SKOS
Names of
metadata
elements
URIs of
classes and
properties
Links to
other URIs
Serialization into RDF
Interoperability 2: RDF and Linked Data
http://purl.org/dc/elements/1.1/contributorhttp://purl.org/ontology/bibo/editor“Editor”
rdfs:subPropertyOf
http://aims.fao.org/aos/agrovoc/c_6599 http://id.loc.gov/authorities/sh85113862#concept“Rice”
skos:exactMatch
6. Names of metadata elements URIs of classes and properties
Databases /
tables / series
Names of things URIs of things Links to other URIs
http://vocabularies.aginfra.eu/soil#isObservedOnLocation
• Then, other complex things like
“URI de-referencing” and “content negotiation”…
(some good triple store platforms do it out of the box)
Interoperability 2: RDF and Linked Data
http://purl.org/ontology/bibo/editor“Editor”
https://aginfra-
sg.ct.infn.it/rdf/cncp/resource/ObservedSoilSite/
16.4CLcch1-1
http://vocabularies.aginfra.eu/soil#Obs
ervedSoilProfile
“Observed soil
16.4CLcch1”
rdfs:type
Serialization into RDF
“Observed in location”
7. Tools used in agINFRA
• For building and managing SKOS:
the FAO VocBench
• For publishing KOSs as Linked Data:
SKOS loaded into Allegrograph
• For building and publishing RDF vocabularies:
Neologism
• For publishing data as Linked Data:
D2RQ from database to RDF
> mapping to published classes and properties
Links are provided in the last slide
8. Linked Data in agINFRA
• Linked Data Vocabularies
– Reference to existing relevant RDF vocabularies and SKOS
– New RDF vocabularies only when not existing (e.g. soil
ontology)
– New KOS only when:
• Not existing
• Mapping needed between local concepts and published concepts
• Extension needed
• Linked Data datasets
– Bibliographic data: AGRIS triple store
– Germplasm data:
• CAAS Linked Data API (presented later)
• CRA triple store (presented later)
– Soil data: CRA triple store (presented later)
10. URLs of agINFRA Linked Data
vocabulary platforms
• agINFRA overview of vocabularies:
http://vocabularies.aginfra.eu
• New agINFRA Soil Vocabulary:
http://vocabularies.aginfra.eu/soil#
• VocBench instances:
http://202.73.13.50:55481/aginfra/
http://artemide.art.uniroma2.it/vocbench2
• Allegrograph triple store of agINFRA KOSs:
http://202.45.139.84:10035/catalogs/fao/repositories/ag
INFRA
11. Namespaces of agINFRA new
vocabularies
• agINFRA Soil vocabulary:
http://vocabularies.aginfra.eu/soil#
• CRA Soil Terms:
http://data.entecra.it/rdf/kos/soil/ or
http://soilmaps.entecra.it/rdf/kos/soil/
• CRA Germplasm Terms:
http://data.entecra.it/rdf/germplasm/soil/ or
http://planta-res.entecra.it/rdf/kos/germplasm/
• agINFRA Resource Types Terms:
http://aginfra.eu/voc/aginfra_doctypes/
• agINFRA Educational Resources Terms:
http://aginfra.eu/voc/aginfra_eduterms/
12. Example 1: the Soil Terms KOS
Rationale:
• CRA had local lists of values for several soil
properties
• In most cases those values mapped
conceptually with terms in published KOSs
Local values published as new KOS with
mappings to USDA Soil Taxonomy terms and/or
WRB whenever possible
15. Example 2: Resource types in AGRIS
Starting from a table
Concept Type of Concept Relationship among concepts
Bibliography Top broadMatch http://purl.org/dc/dcmitype/Text
Book Top NT Handbook/Manual broadMatch http://purl.org/dc/dcmitype/Text
Conference BT Event broadMatch http://purl.org/dc/dcmitype/Event
Dictionary Top broadMatch http://purl.org/dc/dcmitype/Text
Directory Top relatedTerm http://purl.org/dc/dcmitype/Collection
Drawing BT Image broadMatch http://purl.org/dc/dcmitype/Image
Encyclopaedia Top broadMatch http://purl.org/dc/dcmitype/Text
Event Top NT Conference exactMatch http://purl.org/dc/dcmitype/Event
Extension Top
Film Top broadMatch http://purl.org/dc/dcmitype/MovingImage
Graphics BT Image broadMatch http://purl.org/dc/dcmitype/Image
Handbook/Manual BT Book broadMatch http://purl.org/dc/dcmitype/Text
Image Top NT Drawing, NT Graphics, NT MapsorAtlases exactMatch http://purl.org/dc/dcmitype/Image
JournalArticle Top NT Preprint broadMatch http://purl.org/dc/dcmitype/Text
Lit.Review Top broadMatch http://purl.org/dc/dcmitype/Text
Manuscript Top broadMatch http://purl.org/dc/dcmitype/Text
MapsorAtlases BT Image broadMatch http://purl.org/dc/dcmitype/Image
News Top
Non-Conventional Top
NumericalData Top broadMatch http://purl.org/dc/dcmitype/Dataset
Other Top
Patent Top broadMatch http://purl.org/dc/dcmitype/Text
Preprint BT JournalArticle broadMatch http://purl.org/dc/dcmitype/Text
Report Top broadMatch http://purl.org/dc/dcmitype/Text
Sound/Music Top broadMatch http://purl.org/dc/dcmitype/Sound
Speech Top broadMatch http://purl.org/dc/dcmitype/Sound
Standard Top broadMatch http://purl.org/dc/dcmitype/Text
Summary Top broadMatch http://purl.org/dc/dcmitype/Text
Thesaurus Top broadMatch http://purl.org/dc/dcmitype/Text
Thesis Top broadMatch http://purl.org/dc/dcmitype/Text
Website Top
External mapping
19. agINFRA LOD data
19
Germplasm data
http://[CAAS-API-base-URL]/germplasm/rest
https://aginfra-sg.ct.infn.it/rdf/cncp/
CIARD RING
Existing
datasets
CAAS
CRA
New agINFRA datasets
http://ring.ciard.net
DATASETS agINFRA shop? CIARD directories
Germplasm data
Soil data
API
https://aginfra-sg.ct.infn.it/rdf/... ?? CRA Triple store
CRA Triple store
CKAN
CKAN
Dataverse
AGRIS
GLN
20. Namespaces of agINFRA
partners’ Linked Data
• Sustainability namespaces with the data
owners
• CRA data: data.entecra.it (presented later)
– http://data.entecra.it/rdf/soil/
temporarily at https://aginfra-sg.ct.infn.it/rdf/cncp/
– http://data.entecra.it/rdf/germplasm/
temporarily at http://93.63.35.32:8080/d2rq/
• CAAS data (presented later)
– API: http://www.cgris.net/pquery.asp
• AGRIS data
– http://agris.fao.org/aos/
20