Repository navigation
Stream Forge
Stream Forge is the primary running example used throughout Agile Semantic Knowledge Graph Design.
It is a deliberately simplified but realistic domain inspired by large-scale, event-driven content platforms. The example is used to illustrate how semantic knowledge graphs evolve across the full lifecycle—from requirements and conceptual modeling to implementation, governance, and AI-assisted retrieval.
The materials in this section are design artifacts, not tutorials. They are intended to support the book’s discussion of semantic modeling practices rather than to serve as a reference architecture or production system.
Stream Forge is used to demonstrate:
- How requirements can be expressed using stories, competency questions, and lightweight documentation
- How those requirements drive conceptual models defined in OWL
- How conceptual models are refined into design models using SHACL
- How provenance, lineage, and governance metadata are modeled explicitly
- How semantic knowledge graphs can support retrieval-augmented generation (RAG) while making uncertainty and fallback behavior explicit
The domain is intentionally constrained so that the focus remains on modeling decisions, not domain exhaustiveness.
If you are reading the book:
- Treat these pages as supporting artifacts for Chapters 7–15
- You do not need to read everything here to follow the text
- Specific pages may be referenced when additional detail is helpful
If you are browsing independently:
- Start with the Stories section to understand the motivating scenarios
- Then explore the Requirements and Domain Model sections to see how those scenarios are formalized
- Example data, queries, and validation artifacts are provided to make the discussion concrete
Narrative scenarios that motivate requirements and design choices.
- Publishing and Identity
- Metric Provenance and Trust
- Semantic Grounding for AI Responses
Artifacts that capture what the system must support.
- Domain glossary
- Competency questions
- Non-functional considerations
Conceptual descriptions of the core entities and events used in the example.
- Core entities (e.g., videos, creators, platforms)
- Event catalog
Small, readable data samples illustrating ingestion and integration.
- Sample event payloads
- Example metadata extracts
- Example RDF/Turtle fragments
Illustrations of how the modeled knowledge is queried and validated.
- Example SPARQL queries
- SHACL constraint summaries
This section is distinct from the tutorial materials elsewhere in the Wiki:
- Tutorials focus on how to use tools
- Stream Forge focuses on why models are designed the way they are
No step-by-step instructions are provided here. Instead, the emphasis is on reasoning about design trade-offs, constraints, and semantics.
Stream Forge is a living example.
Artifacts may evolve as the book progresses, but changes are made conservatively to preserve consistency with the published text.