Triple level provenance

csv2rdf4lod-automation is licensed under the [Apache License, Version 2.0](https://github.com/timrdf/csv2rdf4lod-automation/wiki/License)

When RDF triples are produced from tabular data, the provenance for each can be captured to cite its source.

The first proof of concept for triple level provenance was the Spring 2010 Advanced Semantic Technologies Project - triple-level provenance - Tech Report, which is also described on inference-web.org.

so if you enable triple level provenance, it spews out NodeSets that justify reifications that omit a subject or predicate.

it omits reifying the subject for each column, and omits reifying the predicate for each row.

so you get M Nodesets justifying predicateless reifications

and N Nodesets justifying subjectless reifications.

Jim McCusker wrote a custom DESCRIBE handler that stitches them together when you query DESCRIBE

Related work

• http://www.bigdata.com/bigdata/blog/