Open Questions concerning data registration

[christian-rli]

1. What level do annotations work on. Is it only on artifact level or can it selectively be done also for a group? In moss I only seem to be able to to annotate specific tables and not an entire group.
2. It looks like we want to treat scenarios as groups, at least that's what seems to make sense for me. In the readme there are [id, label, title, abstract, comment, description] the only values that can be annotated. Do the remaining attributes (I'm thinking of the info listed in the scenario factsheets on the OEP) need to be written to the artifact (table) itself?
3. Can artifacts belong to several groups?
4. Can different versions of artifacts belong to different groups?
5. Is it possible to search for groups in moss? or does one need to look for specific tables first? That would be relevant for a "find scenario" use case
6. I imagine things to work like the following: When registering an artifact there's that field I can use to create random key-value pairs in the python script it's a dictionary e.g. {"emissions": "ghg"}. With that I can create values that show up in the databus view under "Dataid-cv:type" which I can use as a filter (see e.g. here). Is it possible to specifically search for these key value pairs (in moss)?

@0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q @giannou @JJ-Author

[yum-yab]

Answers:

1. No, Annotation level is the Databus File ID, so a singele file in a version of an artifact.
2. Description supports markdown and can be inserted as a markdown table. This will also be indexed by the the databus search engine and be therefor findable on the website.
3. No. For this purpose we recommend using Databus Collections
4. See number 3
5. Not yet, currently the MOSS only returns datasets (identified by Databus File ID).
6. Its currently only possible to search for these content variants (CVs) in SPARQL, there is no implementation in MOSS yet.

[giannou]

Hi @yum-yab, short question on point 6. above: is there an example of how such a SPARQL query looks like? The Databus nicely delivers next to each uploaded file(s) the SPARQL query with which the file(s) can be found, but is there a sample query that contains more abstract things like e.g. metadata?

[henhuy]

Hi @giannou,

in order to test sparqling of metadata, I did following things:

• registered an example data on the databus: https://energy.databus.dbpedia.org/henhuy/DEA/wind_turbine_generation/
• submitted metadata via MOSS to the related DataID (note: metadata contains dummy data! and dummy context.jsonld!) and
• finally created a SPARQL (see below) retriving OEOs for my columns which you can run using YASGUI

SPARQL example:

PREFIX dataid: <http://dataid.dbpedia.org/ns/core#>
PREFIX dct:    <http://purl.org/dc/terms/>
PREFIX dcat:   <http://www.w3.org/ns/dcat#>
PREFIX db:     <https://databus.dbpedia.org/>
PREFIX rdf:    <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX rdfs:   <http://www.w3.org/2000/01/rdf-schema#>
PREFIX csvw: <http://www.w3.org/ns/csvw#>
PREFIX oeo: <http://openenergy-platform.org/ontology/oeo/>
PREFIX obo: <http://purl.obolibrary.org/obo/>
PREFIX prov: <http://www.w3.org/ns/prov#>

SELECT DISTINCT ?label ?description ?unit ?datatype ?about ?value_about WHERE {
  ?dfi csvw:table/csvw:tableSchema ?tableSchema .
  ?tableSchema csvw:column ?col .
  ?col rdfs:label ?label .
  OPTIONAL{?col oeo:OEO_00040010 ?unit .}
  ?col dct:description ?description .
  ?col csvw:datatype ?datatype .
  OPTIONAL{?col obo:IAO_0000136 ?about .}
  ?col obo:IAO_0000010 ?software .
  ?software obo:IAO_0000136 ?value_about .
} 

[giannou]

Thanks a lot, that's very useful