2. Developing Your Ontology7. Exercise Summary
Feedback

7 - Exercise Summary

This content is also available at learn.palantir.com ↗ and is presented here for accessibility purposes.

What you built

  • A simple pipeline developed in Pipeline Builder that outputs an object-backing dataset called flight_alerts_{yourname}_{date}
  • An object type called [yourName] Flight Alert backed by the dataset above
  • A link between your new object type and an existing [Example Data] Flight object type

What you learned

  • Ontology Manager is Foundry's toolkit for building and maintaining your Ontology.
  • A dataset can back a single object type, and an object type can be backed by only one dataset.
  • Backing datasets should be scoped and developed as part of a rigorously maintained data pipeline.
  • After the initial property mapping process, there are still a number of metadata configuration options available to Ontology builders.
  • Link types are more than just formal representations of relationships; they enable cross-object operations in Ontology-aware applications.
  • The object storage service registers a unique table for every object in the ontology and indexes the mapped properties according to the specifications you set in Ontology Manager.
  • Object Explorer is Foundry's application for searching and visualizing object types and Object Views.
  • Data Lineage enables us to visualize the links between object types in the context of a broader data pipeline.