W3C home > Mailing lists > Public > public-lod@w3.org > September 2009

New Version of the Geospecies Knowledge Base Data Set

From: Peter DeVries <pete.devries@gmail.com>
Date: Fri, 4 Sep 2009 19:50:51 -0500
Message-ID: <3833bf630909041750i2ac1c056s841d486a6a156c73@mail.gmail.com>
To: public-lod@w3.org
I have a new version of the GeoSpecies Knowledge Base Dataset at:

http://lod.geospecies.org/

You can download the new data dump at:
http://lod.geospecies.org/geospecies.rdf.tar.tz
It is either in the LOD could or will be soon.

The major differences between this version are:

1) Shorter bit.ly-like identifiers:

Ochlerotatus triseriatus
Old
http://species.geospecies.org/spec_concept_uuid/84b8badd-b899-40ea-8f89-f39f3048c270/
New http://lod.geospecies.org/ses/iuCXz

2) The relationship between entities and documents

At the family and species levels these work in the following way (still need
to port these to RDFa)
The entity itself     http://lod.geospecies.org/ses/iuCXz
The XHTML page  http://lod.geospecies.org/ses/iuCXz.html
The RDF     page  http://lod.geospecies.org/ses/iuCXz.rdf

At the higher levels the pages are xhtml rd.

Order Carnivora
The entity itself               http://lod.geospecies.org/orders/jtSaY
The XHTML RDFa page   http://lod.geospecies.org/orders/jtSaY.html
The RDF page                 http://lod.geospecies.org/orders/jtSaY.rdf

3) The current ontology is in development and I would only use it to look up
the meaning of the different terms.
I am not yet confident that it works as expected with other ontologies. As
with the older version, the ontology
is not included in the RDF data dump.

The ontology is here http://rdf.geospecies.org/ont/geospecies.owl
It is different from the old ontology at:
http://rdf.geospecies.org/ont/gsontology.owl

The ontology documentation is here:
http://lod.geospecies.org/ont_doc/index.html

4) The new vocabulary attempt to use more standard vocabularies including
skos, foaf, and others.

5) My long term intention is to move to this new version and set my old
URI's to redirect to the new URI's.

6) How does this relate to TaxonConcept?
     A) TaxonConcept provides human and machine readable species description
documents that help describe the attributes and variation within a species.
          These will take some time before they become accurate and
informative.

     B) GeoSpecies will eventually be making statements about the entities
defined by TaxonConcept
          What I would like to do is say this entity
http://lod.geospecies.org/ses/iuCXz is defined by
http://lod.taxonconcept.org/ses/iuCXz
          But I would like people to be able to disambiguate the statements
made by geospecies and those by taxonconcept.

          I am not sure the best way to do this and I thought that the
members of this list might be able to give me some good advice.

          Right now they are very similar

          TaxonConcept http://lod.taxonconcept.org/ses/iuCXz
          GeoSpecies     http://lod.geospecies.org/ses/iuCXz

     C) GeoSpecies places the entities into a specific hierarchy which you
may or may  not agree with. The TaxonConcept entities are not
           placed in a hierarchy other than saying that the are species
level entities. The TaxonConcept entities assert that they are a species,
           but not a species in a particular hierarchy. The TaxonConcept
entities can have many different classification hierarchies while at any
           given time the GeoSpecies entities will be one hierarchy which
includes species, family, order, class, phyla, kingdom.

           At this point in time, I don't know if it is best to do this via
types i.e. species X has family type Y, order type Z etc.
           What I have done is create properties inFamily, inOrder etc and
assign a hierarchy that way. I also use skis:broaderTransitive
           and narrower:Transitive between levels of the hierarchy.

           Here is the RDF of the mosquito Ochlerotatus triseriatus
           http://lod.geospecies.org/ses/iuCXz.rdf

           Here is what it looks like in the uriburner

http://linkeddata.uriburner.com/about/html/http/lod.geospecies.org/ses/iuCXz

7) Still to do:
     A) Update the description web pages
     B) Clean up the interlinking relationships and vocabulary .. Any
suggestions
     C) Collect more attribute data for each entity
     D) Add more species
     E) Redo the family and species pages as RDFa
     F) Re-implement observation records, I had tried to port these to
DarwinCore but it has been changing too much.

8) Bugs
    A) The RDF pages are more complete and better tested. I don't have a
working crawler to test the RDFa pages for vocabulary errors and omissions
yet.
    B) The mapping to DBpedia is incomplete there may still be some entities
that map to DBpedia redirects.

I would appreciate any comments or suggestions on how the data set could be
improved. :-)

Respectfully,

- Pete




---------------------------------------------------------------
Pete DeVries <http://spiders.entomology.wisc.edu/pjd/index.html>
Department of Entomology
University of Wisconsin - Madison
445 Russell Laboratories
1630 Linden Drive
Madison, WI 53706
Email: pdevries@wisc.edu
GeoSpecies Knowledge Base <http://species.geospecies.org/>
About the GeoSpecies Knowledge Base <http://about.geospecies.org/>
------------------------------------------------------------
Received on Saturday, 5 September 2009 00:52:39 UTC

This archive was generated by hypermail 2.3.1 : Sunday, 31 March 2013 14:24:23 UTC