tdwg / gbwg

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

SSSOM - split semantic and syntactic predicates #54

Closed pbuttigieg closed 3 years ago

pbuttigieg commented 3 years ago

xref https://github.com/mapping-commons/sssom-py/issues/22

The semantic predicate would focus only on the meaning of each field in the subject/object pair, while the syntactic would note if users can expect a syntax alignment or not. Very useful when writing import functions.

Syntax mismatches can be the result of using different units, value formats, including units in values, etc. - not sure if we want other predicates for each of those cases (may be too much)

In our report, we should recommend closer syntactic alignment or at the very least stability / standard syntax so stable converters can be written.

raissameyer commented 3 years ago

I've now added suggestions for syntax_predicates to all our mappings (see our SSSOM spreadsheet and corresponding issues).

For any measurement values, I've used

tucotuco commented 3 years ago

I don't know if this was covered anywhere else, but would this separation not also suggest that the actual definitions be updated to separate the semantic and syntactic parts? As they are, combined, they often suggest distinct concepts where the "idea" is the same, but the way it is captured is incompatible.

On Fri, Apr 16, 2021 at 9:58 AM Raissa Meyer @.***> wrote:

I've now added suggestions for syntax_predicates to all our mappings (see our SSSOM spreadsheet and corresponding issues https://github.com/tdwg/gbwg/issues?q=is%3Aopen+is%3Aissue+label%3A%22DwC+Mapping%22 ).

For any measurement values, I've used

  • skos:narrowMatch if the mapping was from DwC {float} to MIxS {float}{unit} and both expected the same float (skos:broadMatch for the reverse mapping?)
  • skos:relatedMatch if the mapping was from DwC expecting "verbatim" to MIxS expecting {float}{unit}
  • skos:relatedMatch if the mapping was from DwC expecting a negative value to MIxS expecting a positive value

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/tdwg/gbwg/issues/54#issuecomment-821155504, or unsubscribe https://github.com/notifications/unsubscribe-auth/AADQ724QJT2FYAOT5AEK6LDTJAX6VANCNFSM4ZI6HFUA .

raissameyer commented 3 years ago

Based on the conversations during today's TG meeting, I've updated the comment above https://github.com/tdwg/gbwg/issues/54#issuecomment-821155504 to specify skos:exactMatch for cases where DwC and MIxS both expect numbers (no matter if negative or positive, e.g. see depth)