Return facet list for each page

[anitacaron]

The JSON result should have

{
"data": [{..}],
"pagination": {},
"facets": {
 "mapping_justification": [],
 "predicate_id": [],
 "mapping_provider": [],
 "object_source": [],
 "confidence": [],
 "mapping_set_title": []
}
}

• [x] "mapping_justification"
  • [ ] Return values as curie - blocked by https://github.com/mapping-commons/sssom-py/pull/381
• [x] "predicate_id"
  • [x] Return values as curie
• [ ] "mapping_provider"
  • [ ] Return values as curie
• [ ] "object_source"
  • [ ] Return values as curie
• [x] "confidence"
  • [x] Return min and max values
• [ ] "mapping_set_title"

[anitacaron]

The facets don't have a label for the UI. How to deal with it?

This is how it's returned:

"facets": {
    "mapping_justification": [
      "https://w3id.org/semapv/LogicalReasoning",
      "https://w3id.org/semapv/LexicalMatching"
    ],
    "predicate": [
      "http://www.w3.org/2004/02/skos/core#closeMatch",
      "http://www.w3.org/2004/02/skos/core#exactMatch"
    ],
    "mapping_provider": null,
    "object_source": null
  }

[matentzn]

One option is to have a SEMPAV endpoint embedded in the API. This would allow to do things like returning SEMAPV vocabulary metadata that could be displayed in tooltips in the likes.

[anitacaron]

Can you give a link to the SEMPAV endpoint? I couldn't find it.

[matentzn]

No no, there is no SEMAPV endpoint - only the vocabulary https://github.com/mapping-commons/semantic-mapping-vocabulary/blob/main/semapv.owl

Basically we have two options:

1. We have the UI depend on OLS for obtaining labels for sempav, and make a semapv OLS endpoint mandatory when configuring the UI (this could be ok because we can simply link the public OLS endpoint for this purpose)
2. We add a very basic /terminology endpoint to the SSSOM API to request labels and definitions for semapv / sssom vocabulary. Input: ID (CURIE), output: Label, Definition

I personally prefer the second because this is self contained, but you may want to check with @udp for which solution is advidable.

[anitacaron]

I looked into the vocabulary file semapv.owl, and it seems they have different IRIs used for the mapping_justification values.

The IRI I have when generating SSSOM:

https://w3id.org/semapv/LexicalMatching

The IRI in the semapv.owl file:

https://w3id.org/semapv/vocab/LexicalMatching

Which one is not up to date?

[matentzn]

https://w3id.org/semapv/vocab/LexicalMatching

Is correct.

I had already fixed that in bioregistry: https://bioregistry.io/registry/semapv

Made an issue in sssom py to fix this: https://github.com/mapping-commons/sssom-py/issues/362