Skip to main content

Ontology Lifecycle

The Carbon Arc Ontology is a living framework—designed to evolve as new entities, datasets, and business needs emerge. This document outlines how the Ontology is developed, maintained, and governed across its lifecycle to ensure reliability, transparency, and alignment with real-world use.

Lifecycle Stages​

The lifecycle of every Ontology version follows a repeatable framework:

1. Discovery​

  • Business, product, and data teams identify new real-world entities or representations required to support analytics or platform use cases.
  • Triggers may include: onboarding new datasets, launching new domains, or identifying schema gaps.

2. Definition​

  • New entity types or representations are defined.
  • Mappings between data fields and Ontology nodes are proposed.
  • Semantic standards (e.g., type → representation structure) are applied.

3. Validation​

  • Data science, engineering, and domain teams review changes for:
    • Entity resolution integrity
    • Naming consistency
    • Graph conflicts or duplication
  • Backtests or partial joins are run to validate efficacy.

4. Versioning​

  • A new Ontology version is created.
  • Version identifiers are structured as ontology_vX.Y, e.g. ontology_v2.4.

5. Release​

  • Updates are published across:
    • Ontology API
    • Explorer UI
    • Entity-linked datasets
  • Release notes are posted in our docs

6. Monitoring​

  • Active tracking of issues flagged by users or internal systems.
  • Edge cases (e.g., overlapping representations, deprecations) are added to review backlog.

Version Control​

All Ontology releases are versioned and backward-compatible where possible.

Version IDDescription
v1.xInitial rollout, core entities
v2.xExpanded brand representations
v3.x (planned)Support for product SKUs, metadata tagging

Deprecated representations or structural changes are communicated via changelogs and migration notes.

Governance Model​

Entity Ownership​

Each major Ontology node (e.g., Company, Brand, Location) is stewarded by a domain owner who ensures consistency, hygiene, and alignment across datasets.

Update Cadence​

  • Major versions: Monthly
  • Minor updates: As-needed (monthly or dataset-triggered)

Feedback Loops​

  • Internal requests tracked via Jira
  • External user feedback collected through support and Customer Success
  • Automated detection of mapping conflicts or join failures

Why Ontology Governance Matters​

Without structure, data fragments. Without governance, structure degrades.

A well-managed Ontology ensures:

  • Clean joins between datasets
  • Consistent entity resolution over time
  • Interoperability across platform workflows
  • Trust in downstream analytics and models

For questions or to request a new entity mapping, contact support@carbonarc.co or reach out to your Customer Success Manager.