I am convenor of a group charged with recommending positions about Knowledge Organization Systems (KOS) to GBIF. Among the things that will be in our report will be a survey important vocabularies and ontologies for biodiversity Knowledge Organization. Not only will OBOE be among those discussed, but one of its architects, Mark Schildauer, is a member of the work group. Similarly re: the OBO ontologies. When our report is drafted, it will be opened for public comment both as to what we recommend and what we may have missed. Meanwhile, there is a survey at http://surveymonkey.com/GBIFKOSurvey to which all are invited to participate if you haven't already. If you have favorite KOS resources, there is good opportunity to list them there. Also, the TDWG 2010 Thursday sessions on Observations will, I believe, soon have their notes on the TDWG website (if not there already). Bob Morris On Fri, Oct 22, 2010 at 1:03 AM, Cam Webb <cwebb@oeb.harvard.edu> wrote:

Hi Kevin,

...

I can't help but feel we are getting into a much more complicated area of biodiversity ( or any observation oriented field for that matter ) and that there must be other ontologies or models that we could follow or reuse.

If you are referring specifically to adding depth to the observations of Individuals, I think one might look first at OBOE (OBOE: Extensible Observation Ontology) from the SEEK group at NCEAS:

  http://ecoinformatics.org/oboe/oboe.1.0/oboe.owl

In summary, this describes an Observation of an Entity, with the Observation comprising a Measurement of the Value for a Characteristic of the Entity:

@prefix oboe: <http://ecoinformatics.org/oboe/oboe.1.0/oboe-core.owl#> .

[]  a oboe:Observation ;     oboe:ofEntity [         a oboe:Entity ;         ] ;     oboe:hasMeasurement [         a oboe:Measurement ;         oboe:ofCharacteristic [             a oboe:Characteristic ;             ] ;         oboe:hasValue [             a oboe:Entity ;             ] ;         ] .

If we assert that a dwc:Occurrence is an instance of oboe:Entity, with dwc:basisOfRecord of "HumanObservation", with a dc:creator and dc:created (i.e., the space-time intersection of an Individual and a human observer), we can marry these two ontologies quite nicely.  Please see Example 1, below, for a description of the fruit color of an individual plant.

What's nice about the OBOE model is that it contains all three possible parts of an observation: the entity, the characteristic and the value. This allows direct mapping to the Prometheus Description Model (structure + property + state = entity + characteristic + value; Pullan et al. 2005, Taxon 54:751-765), and indirect mapping to the popular EQ model (entity + quality, where quality = characteristic + value; Mabee et al. 2007, doi:10.1016/j.tree.2007.03.013). However, it doesn't map easily to the DELTA or SDD data models (character + character-state, where character = entity + characteristic and character-state = value).

Because of the well-developed OBO ontologies, we can directly employ terms from, say, the Plant Ontology (po), for an oboe:Entity, and terms from PATO, the quality ontology, for a oboe:Measurement, combining characteristic and value into a quality.  We need to assert that a pato:quality is equivalent to a oboe:Measurement and possibly that a po:PO_0000001 (top level `plant structure') is an oboe:Entity.  We can also employ the OBO relational ontology (ro) to indicate that a fruit is ro:part_of the particular space-time Occurrence of an Individual (this might require a bit more discussion!).

So, it seems that using only well-established vocabularies, we can make semantic statements about the characteristics of individuals originally defined primarily using DwC terms.  Please see example 2, below for a fairly slim, usable model of a description of the fruit color of an individual plant.  The image of the model is at:

  http://phylodiversity.net/cwebb/img/tdwg-obs.jpg

Again, any comments on this model will be much appreciated.  Are you aware of other attempts to join DwC Occurrence models with OBOE models?

Best,

Cam

====================== Example 1 =======================================

@prefix oboe: <http://ecoinformatics.org/oboe/oboe.1.0/oboe-core.owl#> . @prefix dwc: <http://rs.tdwg.org/dwc/terms/> . @prefix dcterms: <http://purl.org/dc/terms/> . @prefix sernec: <http://bioimages.vanderbilt.edu/rdf/terms#> . @prefix geo: <http://www.w3.org/2003/01/geo/wgs84_pos#> . @prefix : <#> .

<http://phylodiversity.net/xmalesia/indiv/9>     a sernec:Individual ;     sernec:derivativeOccurrence _:blank1 .

_:blank1     # The Occurrence of the Individual...     a dwc:Occurrence ;     # ... at a position in space-time...     dcterms:created "2008-01-01" ;     dcterms:spatial [         a dcterms:Location ;         geo:lon "109.95371" ;         geo:lat "-1.25530" ;         dwc:coordinateUncertaintyInMeters "100" ;         ] ;     # is a recordable OBOE Entity        a oboe:Entity ;     # as recorded by a human     dcterms:creator "Cam Webb" ;     dwc:basisOfRecord "HumanObservation" .

# The details of the observation: []  a oboe:Observation ;     oboe:ofEntity [         # The observed entity is actually *part of* the occurrence         #   of the Individual at a particular Space and Time         a :Fruit ;         :partOf _:blank1 ;         ] ;     oboe:hasMeasurement [         oboe:ofCharacteristic :Color ;         oboe:hasValue :Green ;         ] .

:Fruit a oboe:Entity . :Color a oboe:Characteristic . :Green a oboe:Entity .

========================================================================

====================== Example 2 =======================================

@prefix oboe: <http://ecoinformatics.org/oboe/oboe.1.0/oboe-core.owl#> . @prefix dwc: <http://rs.tdwg.org/dwc/terms/> . @prefix dcterms: <http://purl.org/dc/terms/> . @prefix sernec: <http://bioimages.vanderbilt.edu/rdf/terms#> . @prefix geo: <http://www.w3.org/2003/01/geo/wgs84_pos#> . @prefix ro: <http://www.obofoundry.org/ro/ro.owl#> . @prefix pato: <http://purl.org/obo/owl/PATO#> . @prefix po: <http://purl.org/obo/owl/PO#> . @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .

<http://phylodiversity.net/xmalesia/indiv/9>     sernec:derivativeOccurrence _:blank1 .

_:blank1     a dwc:Occurrence ;     dcterms:created "2008-01-01" ;     dcterms:spatial [         geo:lon "109.95371" ;         geo:lat "-1.25530" ;         dwc:coordinateUncertaintyInMeters "100" ;         ] ;     dcterms:creator "Cam Webb" ;     dwc:basisOfRecord "HumanObservation" .

# The details of the observation: []  a oboe:Observation ;     oboe:ofEntity [         ro:part_of _:blank1 ;         a po:PO_0009001 ; # Fruit         ] ;     oboe:hasMeasurement pato:PATO_0000320 . # Green color

po:PO_0009001 rdfs:label "fruit" . pato:PATO_0000320 rdfs:label "green" .

======================================================================== _______________________________________________ tdwg-content mailing list tdwg-content@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/tdwg-content

-- Robert A. Morris Emeritus Professor  of Computer Science UMASS-Boston 100 Morrissey Blvd Boston, MA 02125-3390 Associate, Harvard University Herbaria email: morris.bob@gmail.com web: http://bdei.cs.umb.edu/ web: http://etaxonomy.org/mw/FilteredPush http://www.cs.umb.edu/~ram phone (+1) 857 222 7992 (mobile)

People should be marking up these vocabularies with real data and demonstrating how they can be used to make useful queries in a viewable sparql endpoint or the LOD cloud endpoint. Also demonstrating how the interoperate with related datasets. - Pete On Fri, Oct 22, 2010 at 12:03 AM, Cam Webb <cwebb@oeb.harvard.edu> wrote:

Hi Kevin,

...

I can't help but feel we are getting into a much more complicated area of biodiversity ( or any observation oriented field for that matter ) and that there must be other ontologies or models that we could follow or reuse.

If you are referring specifically to adding depth to the observations of Individuals, I think one might look first at OBOE (OBOE: Extensible Observation Ontology) from the SEEK group at NCEAS:

http://ecoinformatics.org/oboe/oboe.1.0/oboe.owl

In summary, this describes an Observation of an Entity, with the Observation comprising a Measurement of the Value for a Characteristic of the Entity:

@prefix oboe: <http://ecoinformatics.org/oboe/oboe.1.0/oboe-core.owl#> .

[] a oboe:Observation ; oboe:ofEntity [ a oboe:Entity ; ] ; oboe:hasMeasurement [ a oboe:Measurement ; oboe:ofCharacteristic [ a oboe:Characteristic ; ] ; oboe:hasValue [ a oboe:Entity ; ] ; ] .

If we assert that a dwc:Occurrence is an instance of oboe:Entity, with dwc:basisOfRecord of "HumanObservation", with a dc:creator and dc:created (i.e., the space-time intersection of an Individual and a human observer), we can marry these two ontologies quite nicely. Please see Example 1, below, for a description of the fruit color of an individual plant.

What's nice about the OBOE model is that it contains all three possible parts of an observation: the entity, the characteristic and the value. This allows direct mapping to the Prometheus Description Model (structure + property + state = entity + characteristic + value; Pullan et al. 2005, Taxon 54:751-765), and indirect mapping to the popular EQ model (entity + quality, where quality = characteristic + value; Mabee et al. 2007, doi:10.1016/j.tree.2007.03.013). However, it doesn't map easily to the DELTA or SDD data models (character + character-state, where character = entity + characteristic and character-state = value).

Because of the well-developed OBO ontologies, we can directly employ terms from, say, the Plant Ontology (po), for an oboe:Entity, and terms from PATO, the quality ontology, for a oboe:Measurement, combining characteristic and value into a quality. We need to assert that a pato:quality is equivalent to a oboe:Measurement and possibly that a po:PO_0000001 (top level `plant structure') is an oboe:Entity. We can also employ the OBO relational ontology (ro) to indicate that a fruit is ro:part_of the particular space-time Occurrence of an Individual (this might require a bit more discussion!).

So, it seems that using only well-established vocabularies, we can make semantic statements about the characteristics of individuals originally defined primarily using DwC terms. Please see example 2, below for a fairly slim, usable model of a description of the fruit color of an individual plant. The image of the model is at:

http://phylodiversity.net/cwebb/img/tdwg-obs.jpg

Again, any comments on this model will be much appreciated. Are you aware of other attempts to join DwC Occurrence models with OBOE models?

Best,

Cam

====================== Example 1 =======================================

@prefix oboe: <http://ecoinformatics.org/oboe/oboe.1.0/oboe-core.owl#> . @prefix dwc: <http://rs.tdwg.org/dwc/terms/> . @prefix dcterms: <http://purl.org/dc/terms/> . @prefix sernec: <http://bioimages.vanderbilt.edu/rdf/terms#> . @prefix geo: <http://www.w3.org/2003/01/geo/wgs84_pos#> . @prefix : <#> .

<http://phylodiversity.net/xmalesia/indiv/9> a sernec:Individual ; sernec:derivativeOccurrence _:blank1 .

_:blank1 # The Occurrence of the Individual... a dwc:Occurrence ; # ... at a position in space-time... dcterms:created "2008-01-01" ; dcterms:spatial [ a dcterms:Location ; geo:lon "109.95371" ; geo:lat "-1.25530" ; dwc:coordinateUncertaintyInMeters "100" ; ] ; # is a recordable OBOE Entity a oboe:Entity ; # as recorded by a human dcterms:creator "Cam Webb" ; dwc:basisOfRecord "HumanObservation" .

# The details of the observation: [] a oboe:Observation ; oboe:ofEntity [ # The observed entity is actually *part of* the occurrence # of the Individual at a particular Space and Time a :Fruit ; :partOf _:blank1 ; ] ; oboe:hasMeasurement [ oboe:ofCharacteristic :Color ; oboe:hasValue :Green ; ] .

:Fruit a oboe:Entity . :Color a oboe:Characteristic . :Green a oboe:Entity .

========================================================================

====================== Example 2 =======================================

@prefix oboe: <http://ecoinformatics.org/oboe/oboe.1.0/oboe-core.owl#> . @prefix dwc: <http://rs.tdwg.org/dwc/terms/> . @prefix dcterms: <http://purl.org/dc/terms/> . @prefix sernec: <http://bioimages.vanderbilt.edu/rdf/terms#> . @prefix geo: <http://www.w3.org/2003/01/geo/wgs84_pos#> . @prefix ro: <http://www.obofoundry.org/ro/ro.owl#> . @prefix pato: <http://purl.org/obo/owl/PATO#> . @prefix po: <http://purl.org/obo/owl/PO#> . @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .

<http://phylodiversity.net/xmalesia/indiv/9> sernec:derivativeOccurrence _:blank1 .

_:blank1 a dwc:Occurrence ; dcterms:created "2008-01-01" ; dcterms:spatial [ geo:lon "109.95371" ; geo:lat "-1.25530" ; dwc:coordinateUncertaintyInMeters "100" ; ] ; dcterms:creator "Cam Webb" ; dwc:basisOfRecord "HumanObservation" .

# The details of the observation: [] a oboe:Observation ; oboe:ofEntity [ ro:part_of _:blank1 ; a po:PO_0009001 ; # Fruit ] ; oboe:hasMeasurement pato:PATO_0000320 . # Green color

po:PO_0009001 rdfs:label "fruit" . pato:PATO_0000320 rdfs:label "green" .

======================================================================== _______________________________________________ tdwg-content mailing list tdwg-content@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/tdwg-content

-- ---------------------------------------------------------------- Pete DeVries Department of Entomology University of Wisconsin - Madison 445 Russell Laboratories 1630 Linden Drive Madison, WI 53706 TaxonConcept Knowledge Base <http://www.taxonconcept.org/> / GeoSpecies Knowledge Base <http://lod.geospecies.org/> About the GeoSpecies Knowledge Base <http://about.geospecies.org/> ------------------------------------------------------------

Dear Hilmar, Thanks for this insight. I have to admit to no prior experience of mereology, but the desire to find a solution to how to semantically encode morphological observations in a fashion which many can agree upon prods me to reply. Sorry if I am way in out of my depth and it shows!

...

We can also employ the OBO relational ontology (ro) to indicate that a fruit is ro:part_of the particular space-time Occurrence of an Individual (this might require a bit more discussion!).

I don't think you can do that actually, given the definition of the relation [1]. I.e., to me the definition sounds like it is true meristic parthood that is meant here (and for which it is applied in the uses that I have seen). The fruit has some role in Occurrence, but isn't a part of it in the sense of meristics, don't you think?

If we consider the occurrence to be the intersection of the individual with an (observation) event in a particular space and time (e.g., tzxy0), I think a fruit_txyz0 is indeed part of plant_txyz0, fulfilling the ro:part_of definition of: ``For continuants: C part_of C' if and only if: given any c that instantiates C at a time t, there is some c' such that c' instantiates C' at time t, and c *part_of* c' at t.'' The alternative to making the fruit a part of the dwc:Occurrence is to make it a part_of the sernec:Individual, which could then be observed via the OBOE terms, at a particular place and time (although strangely there seems to be no date/time observation terms in OBOE). This second alternative would be less satisfying though, given that dwc:Occurrences have already been specified to include Observations. Maybe using dcterms:isPartOf would be OK, since it is less precisely defined?

Also, the quantification doesn't match: If A ro:part_of B, then all instances of A part_of some instance of B. Obviously, there are some fruits that haven't been recorded in an occurrence.

I agree with this, but this must surely be a generic issue within the whole OBO/organismal annotation world - often a phenotype must be recorded for c, which is part_of c', without having considered every instance of C. This is clearly an issue to take up with the TDWG Observations WG, and I'll post to the new post-Wood's Hole list shortly. If any of you have any suggestions for how to progress with this, I'd appreciate them. Best, Cam

Hi Hilmar,

...

I agree with this, but this must surely be a generic issue within the whole OBO/organismal annotation world - often a phenotype must be recorded for c, which is part_of c', without having considered every instance of C.

In the EQ model, the phenotype is recorded as some quality Q that inheres in some entity E. E may be a composition (OWL intersection), for example E that is part_of some E' (e.g., the process that is part_of the urohyal bone): some quality Q that inheres in some (E and part_of some E'). There is no need to consider every E: the statement asserts that there is some instance of E that is also part_of some instance of E' and in which some instance of Q inheres.

Could you suggest how to specify in RDF, using EQ terms, how I'd say ``the subset that represents a majority of instances of all fruits that are part_of Individual X'' I'm having trouble with this one! Thanks. Cam