Semantic extensions to ecoRelevé
Olivier Rovellotti
• The projet
• The Sem Web
• The extentions
Reduce
Software complexity
For who?
People we know:




                  People we don’t know?
What for?

                 Ask        Analyze
               Questions     data




                Define
  Data Admin   Protocols   Collect data
DBA/GIS   Research
Professional




Citizen Science                 PHD




                  3 Modules
RDF
2007
2009
2012




Life Science
Now what ?
What is it good for?



     •   Better annotations
     •   Easier data integration
     •   More extensible
     •   More expressive
Aggregate data
  Controlled
 Vocabularies:                     Enhance existing
Build protocols                        dataset
     forms
Annotations
ecoRelevé
Core
1.0



Data integration




        Observation   Layers
Data enhancement

        ecoRelevé
        explorer
4 dimensions
                        Space




               People           Biology


                        Time
GeoSparql




                 Localities                               Protected Areas


       Work                                                       Conservation Status


                                  Space


FOAF                                   Time
                         People               Biology
       Friends                                                    Family




                 At the same time as          Last week
Data connector
Milan royal
Milvus milvus
Red Kite
Rod Page - what can you do with it ?
SPARQL:
Catch the frog
                                                      TAISTY!


select ?scientificName, ?status, ?lat, ?long                    ecoReleve
FROM <urn:rdf.TdwgFroggyChallenge>
where {
 ?s rdf:type uniprot:Molecule .
 ?s terms:relation ?id.
 ?s terms:subject ?taxonInGB .
                                                                      SparQL
 ?id geo:lat ?lat.
 ?id geo:long ?long.

 ?taxonInDbPedia dbOwl:conservationStatus ?status.
 ?taxonInGB rdfs:seeAlso ?taxonInDbPedia.
 ?taxonInGB uniprot:scientificName ?scientificName.
}
Take home message :


    •   There is a learning curve
    •   But RDF is not that difficult
    •   One API is better than 10
    •   We need data in RDF to experiment

    • Reasoning is for later …
Thanks ..


@orovellotti

Tdwg2011