Documentation / Concepts
Graph Semantics
The decision graph is not only a structure of nodes and edges. It is a semantic model where entities, relationships and signals carry meaning in context.
Graph semantics define what the graph means, not only how it can be traversed. In decision systems, meaning matters because a result is only useful if the surrounding context remains interpretable.
Meaning in a Graph
A graph becomes semantically useful when its nodes, edges and attached signals correspond to meaningful parts of the decision environment. The graph is therefore not just a storage pattern, but a structured representation of context.
Semantic graph view
A conceptual graph where entities, relationships and attached signals remain distinct but connected.
Entities
Entities are the durable subjects of reasoning inside the graph. They represent the parts of the environment that persist long enough to carry context between decisions.
organisations
suppliers
contracts
systems
events
Relationships
Relationships preserve how entities depend on, constrain or influence one another. In a decision graph, relationships are not incidental links. They are part of the semantics of the environment itself.
dependencies
ownership paths
governing constraints
temporal or operational links
Signals
Signals attach evidence, state or inferred meaning to entities and relationships. They provide the changing informational layer that decision systems must evaluate without losing provenance or type.
asserted facts from source systems
probabilistic assessments from model outputs
derived signals created through reasoning
Context
Context
Context emerges from the combination of entities, relationships and attached signals rather than from any single record viewed in isolation.
In semantic graph reasoning, the meaning of a node depends on the surrounding structure. The same entity can support different decision paths depending on what relationships and signals are currently attached to it.
Why Semantics Matter
If a graph preserves structure but not meaning, it becomes difficult to explain why a decision path exists or what a connected signal actually represents.
Semantic structure is what allows the decision graph to support traceable reasoning rather than generic graph traversal. It turns the graph into an interpretable model of the decision environment.