- From: Stasinos Konstantopoulos <konstant@iit.demokritos.gr>
- Date: Sat, 18 Oct 2014 11:20:36 +0300
- To: Dan Brickley <danbri@google.com>
- Cc: "public-csv-wg@w3.org" <public-csv-wg@w3.org>
Dan, all, there's also dbpedia, which comes as a set of 685 CSVs in addition to RDF [1]. So that gives us both the CSV and the RDF that is intended as equivalent. I will describe it in more detail, if you find this use case interesting. Best, s [1] http://wiki.dbpedia.org/DBpediaAsTables On 16 October 2014 17:37, Dan Brickley <danbri@google.com> wrote: > On 16 October 2014 12:49, Anastasia Dimou <anastasia.dimou@ugent.be> wrote: >> >> On 10/15/2014 03:44 PM, Dan Brickley wrote: >>> >>> http://chinookdatabase.codeplex.com/ >>> >>> I've uploaded a shell script that converts from the Sqlite3 dump into >>> CSV, alongside the CSVs. >>> >>> See >>> https://github.com/w3c/csvw/tree/gh-pages/examples/tests/scenarios/chinook >>> for overview of schema / contents and links to the data files, >>> >>> https://github.com/w3c/csvw/tree/gh-pages/examples/tests/scenarios/chinook/csv >>> >>> Intent is to use this as a well known dataset for exploring multi-CSV >>> issues with mappings/templates. >> >> >> And I uploaded an RML mapping document mapping the chinook CSV files to RDF. >> See >> https://github.com/w3c/csvw/tree/gh-pages/examples/tests/scenarios/chinook/attempts/attempt-1. >> >> Remark: I changed the public RMLProcessor (development branch) to make it >> even simpler to run the mappings just by providing the mapping document (URL >> or local file) and the output file, for instance mvn exec:java >> -Dexec.args="https://raw.githubusercontent.com/w3c/csvw/gh-pages/examples/tests/scenarios/chinook/attempts/attempt-1/chinook.rml.ttl >> /path/to/local/output/file/chinook.rml.nt" > > This is super great - thanks Anastasia! We now have an example of at > least one way of mapping a package of multiple CSVs into a common > graph. > > Are there particular bundles of CSVs from other use cases we should > look at next? > > The events demo (single csv) is now updated to use the new simpler > tooling / interface. > > https://github.com/w3c/csvw/tree/gh-pages/examples/tests/scenarios/events > > >From https://github.com/w3c/csvw/blob/gh-pages/examples/tests/scenarios/events/attempts/attempt-1/mapping-events.rml.ttl > here is the mapping. > > <#myCSV> rml:source > "https://raw.githubusercontent.com/w3c/csvw/gh-pages/examples/tests/scenarios/events/source/events-listing.csv"; > rml:referenceFormulation ql:CSV . > <#MusicEvent> > rml:logicalSource <#myCSV>; > rr:subjectMap [ rr:termType rr:BlankNode; rr:class schema:MusicEvent; ]; > rr:predicateObjectMap [ rr:predicate schema:name; rr:objectMap [ > rml:reference "Name"; ] ]; > rr:predicateObjectMap [ rr:predicate schema:startDate; rr:objectMap [ > rml:reference "StartDate"; rr:datatype schema:Date; ] ]; > rr:predicateObjectMap [ rr:predicate schema:location; rr:objectMap [ > rr:parentTriplesMap <#Place> ] ]; > rr:predicateObjectMap [ rr:predicate schema:offers; rr:objectMap [ > rr:parentTriplesMap <#Offer> ] ] . > <#Place> > rml:logicalSource <#myCSV>; > rr:subjectMap [ rr:termType rr:BlankNode; rr:class schema:Place; ]; > rr:predicateObjectMap [ rr:predicate schema:address; rr:objectMap [ > rml:reference "location_address"; ] ]; > rr:predicateObjectMap [ rr:predicate schema:name; rr:objectMap [ > rml:reference "location_name"; ] ] . > <#Offer> > rml:logicalSource <#myCSV>; > rr:subjectMap [ rr:termType rr:BlankNode; rr:class schema:Offer; ]; > rr:predicateObjectMap [ rr:predicate schema:url; rr:objectMap [ > rml:reference "ticket_url"; rr:termType rr:IRI; ] ] . > > > > The next natural question is: how should this be cited from the CSV's > JSON-LD metadata description? > > Can anyone help make > https://github.com/w3c/csvw/blob/gh-pages/examples/tests/scenarios/events/attempts/attempt-1/metadata.json > look more plausible? > > Dan >
Received on Saturday, 18 October 2014 08:21:13 UTC