Step: Publish to Knowledge Base

This step will publish the contents of the _default_ graph to Working Memory, Contextual Memory, or another RDF Data Source you have configured.

Operates on a reified package graph

Choose the location to save the graph to

  • Working Memory: The part of the database that DarkLight uses to store published incoming data.
  • Contextual Memory: Content ingested into DarkLight to provide context for your PROs. This is data like employee records, property databases, etc.
  • RDF Data Source: Any Data Source you create with a type of RDF will also be available for publishing

Choose to merge or replace any matching graphs

Each graph in the database is defined by it's IRI. Typically, the IRI for the graph is created by the reify process and can either be generated, or created from a value in the incoming data. An IRI can also be created in a playbook with the Create New Object step. Graphs with multiple objects in them have a Primary Publish Object who's IRI is used to identify the graph.

If there is already a graph in the database that has the exact IRI of the graph being published, this choice will determine if the graph and all of its connections should be combined with the existing one, or completely replace the existing one.

A good example of a need to replace a graph is when you ingest Employee records and you want to make sure that any properties deleted from the source are also deleted from the stored object.

An example of a merge might be where you have created a new object and are attaching incoming events that match certain criteria in your playbook. Each time the playbook publishes to this object, it will not remove any of the previously-published objects or properties.

  • step/publish
  • Last modified: 2019/03/29 22:57