tdwg / gbwg

Genomic Biodiversity Interest Group
Apache License 2.0
13 stars 2 forks source link

Review and update skos:narrowMatch and skos:broadMatch mappings #77

Closed raissameyer closed 1 year ago

raissameyer commented 2 years ago

This PR and issue in the SSSOM repo showed that - by following the old SSSOM guidance - we have sometimes used the SKOS predicates for narrow match and broad match incorrectly.

Definition in old SSSOM guidance:

skos:narrowMatch: The subject is taxonomically narrower than the object. skos:broadMatch: The subject is taxonomically broader than the object.

The updated definitions are in accordance with the SKOS specs:

skos:narrowMatch: From the SKOS primer: A triple \ skos:narrower (and skos:narrowMatch) \ asserts that \, the object of the triple, is a narrower concept than \, the subject of the triple skos:broadMatch: From the SKOS primer: A triple \ skos:broader (and skos:broadMatch) \ asserts that \, the object of the triple, is a broader concept than \, the subject of the triple

The SSSOM guidance has since been updated.

In this issue we will review and update our mapping files to fix this.


Please see below for a list of the mappings that use skos:narrowMatch or skos:broadMatch and may require updating:

Semantic mappings

V1 skos:broadMatch, potentially need to be update to skos:narrowMatch

  • dwc:verbatimCoordinates [Verbatim Coordinates] to MIXS:0000009 [lat_lon]
  • dwc:country [Country] to MIXS:0000010 [geo_loc_name]
  • dwc:waterBody [Water Body] to MIXS:0000010 [geo_loc_name]
  • dwc:associatedOrganisms [Associated Organisms] to MIXS:0000028 [biotic_relationship]
  • dwc:samplingProtocol [Sampling Protocol] to MIXS:0000002 [samp_collect_device] update comment!
  • dwc:sampleSizeValue [Sample Size Value] to MIXS:0000001 [samp_size]
  • dwc:sampleSizeUnit [Sample Size Unit] to MIXS:0000001 [samp_size]

V1 skos:narrowMatch, potentially need to be updated to skos:broadMatch

  • dwc:decimalLatitude [Decimal Latitude] toMIXS:0000009 [lat_lon]
  • dwc:decimalLongitude [Decimal Longitude] to MIXS:0000009 [lat_lon]
  • dwc:verbatimLatitude [Verbatim Latitude] to MIXS:0000009 [lat_lon]
  • dwc:verbatimLongitude [Verbatim Longitude] to MIXS:0000009 [lat_lon]
  • dwc:verbatimDepth [Verbatim Depth] to MIXS:0000018 [depth]
  • dwc:minimumDistanceAboveSurfaceInMeters [Minimum Distance Above Surface in Meters] to MIXS:0000018 [depth]
  • dwc:maximumDistanceAboveSurfaceInMeters [Maximum Distance Above Surface in Meters] to MIXS:0000018 [depth]
  • dwc:minimumDepthInMeters [Minimum Depth In Meters] to MIXS:0000018 [depth]
  • dwc:maximumDepthInMeters [Maximum Depth In Meters] to MIXS:0000018 [depth]
  • dwc:minimumDistanceAboveSurfaceInMeters [Minimum Distance Above Surface in Meters] to [alt]
  • dwc:maximumDistanceAboveSurfaceInMeters [Maximum Distance Above Surface in Meters] to [alt]
  • dwc:minimumElevationInMeters [Minimum Elevation in Meters] to MIXS:0000093 [elev]
  • dwc:maximumElevationInMeters [Maximum Elevation in Meters] to MIXS:0000093 [elev]
  • dwc:eventDate [Event Date] to MIXS:0000011 [collection_date]
  • dwc:samplingProtocol [Sampling Protocol] to MIXS:0000016 [samp_mat_process]
  • dwc:associatedSequences [Associated Sequences] to MIXS:0000091 [url]
  • dwc:associatedTaxa [Associated Taxa] to MIXS:0000091 [url]

Syntactic mappings

V1 skos:narrowMatch, potentially need to be updated to skos:broadMatch

  • dwc:decimalLatitude [Decimal Latitude] to MIXS:0000009 [lat_lon]
  • dwc:decimalLongitude [Decimal Longitude] to MIXS:0000009 [lat_lon]
  • dwc:minimumDepthInMeters [Minimum Depth In Meters] to MIXS:0000018 [depth]
  • dwc:maximumDepthInMeters [Maximum Depth In Meters] to MIXS:0000018 [depth]
  • dwc:minimumDistanceAboveSurfaceInMeters [Minimum Distance Above Surface in Meters] to [alt]
  • dwc:maximumDistanceAboveSurfaceInMeters [Maximum Distance Above Surface in Meters] to [alt]
  • dwc:minimumElevationInMeters [Minimum Elevation in Meters] to MIXS:0000093 [elev]
  • dwc:maximumElevationInMeters [Maximum Elevation in Meters] to MIXS:0000093 [elev]
raissameyer commented 2 years ago

List of terms with updated semantic mappings

skos:broadMatch updated to skos:narrowMatch "subject has skos:narrowMatch object"

  • dwc:verbatimCoordinates [Verbatim Coordinates] skos:narrowMatch MIXS:0000009 [lat_lon]
  • dwc:samplingProtocol [Sampling Protocol] skos:narrowMatch MIXS:0000002 [samp_collect_device]
    • xRef https://github.com/tdwg/gbwg/issues/20#issue-805169799
    • Update comment: dwc:samplingProtocol covers both of the concepts samp_collect_device and samp_collect_process. Both of these terms list both, the names of (or links to) methods, or devices (either in the definition or in the examples). Based on the examples these could be skos:exactMatch, focusing on the definition, skos:narrowMatch would be appropriate. The information that this term and the samp_mat_process capture is often captured in the method steps in the EML (metadata about the dataset, not indexed in GBIF’s case), and less in the DwC terms themselves. This influences how people will use the terms in reality.
raissameyer commented 2 years ago

List of terms with updated semantic mappings

skos:narrowMatch updated to skos:broadMatch "subject has skos:broadMatch object"

raissameyer commented 2 years ago

List of terms with updated syntactic mappings

skos:narrowMatch updated to skos:broadMatch "subject has skos:broadMatch object"

raissameyer commented 2 years ago

While reviewing the mapping, the following mappings were additionally updated:

syntax xRef https://github.com/tdwg/gbwg/issues/13 (previously skos:exactMatch)

  • dwc:minimumDistanceAboveSurfaceInMeters skos:broadMatch MIXS:0000018[depth] (the DwC term only expects the value, while the MIxS term expects both value and unit)
  • dwc:maximumDistanceAboveSurfaceInMeters skos:broadMatch MIXS:0000018[depth] (the DwC term only expects the value, while the MIxS term expects both value and unit)