When a SPARQL micro-service is invoked, it generates a small, resource-centric graph delineated by the Web API service being wrapped, the arguments passed to the service, and the types of RDF triples that the SPARQL micro-service is designed to spawn.
Optionally, provenance information can be added to the graph. This is configured on a micro-service--basis with parameter add_provenance = truein the config.ini file, or with triple <> sms:addProvenance "true"^^xsd:boolean; in the service description graph (see the configuration page).
Two SPARQL queries are used to generate the provenance information:
- add_provenance.sparql applies to micro-services configured using the service description method
- add_provenance_simple.sparql applies to micro-services configured using the config.ini method, and generates more simple provenance information (because some information, such as the Web API name and URL, are only provided in the service description method).
The provenance information is modelled according to the Provenance Ontology (PROV-O) as described below.
- The top blank node (
_:b1) is an Entity representing the graph produced. It was generated at a certain date and time, and generated by an Activity (_:b2) denoting the invocation of the SPARQL micro-service. - The SPARQL micro-service invocation (
_:b2) was associated with a SoftwareAgent (_:b3) that is the SPARQL micro-service itself, i.e. the micro-service was responsible for executing the invocation activity. Plus, the SPARQL micro-service invocation (_:b2) used as an input an Entity (_:b4) denoting the result of the Web API invocation. - The Web API invocation result (
_:b4) was generated at a certain date and time, and generated by an Activity (_:b5) denoting the Web API invocation. - The API invocation (
_:b5) was associated with a SoftwareAgent (_:b6) that is the Web API, and was invoked by means of an HTTP request URI.
Note that the graph (_:b1) and the Web API invocation result (_:b4) may have been generated at different dates and times. This happens when a cache database is used: the Web API invocation result was generated at the date and time at which the cached document was fetched from the WebAPI. Whereas the graph was generated at the date and time the SPARQL micro-service was invoked. Hence, the Web API invocation may be older than the SPARQL micro-service invocation.