The ontology maps the DataCite relation type Collects to cito:usesDataFrom, as shown in the SPAR Datacite ontology documentation. However, the two properties do not align semantically, particularly with respect to the expected domain and direction of the relationship.
In the DataCite schema, Collects describes a relationship where resource A (typically an instrument or facility) was used to collect or measure the data in resource B. The subject is therefore a physical tool or system, and the object is the dataset it produces. Example: Radio telescope → observed dataset.
In CiTO, usesDataFrom is a sub-property of cito:cites whose implicit domain is a citing entity — a publication, report, software, etc. — that makes use of data originating from another resource. The subject is thus an intellectual or technical work, not a physical instrument. Example: Paper → dataset analyzed in the paper.
Applying usesDataFrom to the Collects case inverts the intended direction: one would end up stating that the telescope uses data from the dataset, whereas the correct reading is that the dataset originates from the telescope. There is no exact CiTO counterpart for the "instrument → produced data" relationship, which may explain why the nearest available property was chosen.
Note also that the reverse relation (IsCollectBy mapped to cito:providesDataFor) has the same issue.
I suggest either:
-
Adding a clarifying annotation on the mapping noting that it is approximate and does not preserve the original semantics of Collects, or
-
Reconsidering the target property, but I didn't find any equivalent property yet, to better reflect the instrument-to-data relationship. I thought of dcterms:source, which could rather be applicable to IsCollectBy.
Happy to discuss further.
The ontology maps the DataCite relation type
Collectstocito:usesDataFrom, as shown in the SPAR Datacite ontology documentation. However, the two properties do not align semantically, particularly with respect to the expected domain and direction of the relationship.In the DataCite schema,
Collectsdescribes a relationship where resource A (typically an instrument or facility) was used to collect or measure the data in resource B. The subject is therefore a physical tool or system, and the object is the dataset it produces. Example: Radio telescope → observed dataset.In CiTO,
usesDataFromis a sub-property ofcito:citeswhose implicit domain is a citing entity — a publication, report, software, etc. — that makes use of data originating from another resource. The subject is thus an intellectual or technical work, not a physical instrument. Example: Paper → dataset analyzed in the paper.Applying
usesDataFromto theCollectscase inverts the intended direction: one would end up stating that the telescope uses data from the dataset, whereas the correct reading is that the dataset originates from the telescope. There is no exact CiTO counterpart for the "instrument → produced data" relationship, which may explain why the nearest available property was chosen.Note also that the reverse relation (
IsCollectBymapped tocito:providesDataFor) has the same issue.I suggest either:
Adding a clarifying annotation on the mapping noting that it is approximate and does not preserve the original semantics of
Collects, orReconsidering the target property, but I didn't find any equivalent property yet, to better reflect the instrument-to-data relationship. I thought of
dcterms:source, which could rather be applicable toIsCollectBy.Happy to discuss further.