Glossary · How AI reads content
Knowledge Graph
A structured network of entities — people, places, brands, concepts — and the relationships between them, helping machines understand meaning rather than keywords.
Entities and their relationships
A knowledge graph is a structured network of entities — people, places, brands, concepts — and the relationships that connect them. It’s how a machine moves from matching text to actually understanding that a word refers to a specific thing in the world.
Things, not strings
When Google introduced its Knowledge Graph in 2012, it framed the shift with a phrase that still captures the idea: understanding “things, not strings.” A traditional index treats a query as characters to match against pages. A knowledge graph treats it as a reference to an entity— a distinct thing with its own attributes and connections — so the system can reason about it rather than just find the letters.
Each entity is a node; each relationship is an edge. “Founded in,” “based in,” “competes with,” “category” — these links are what let a machine answer a question about a thing by traversing what it’s connected to.
Why it matters
The graph’s most visible output is the knowledge panel — the box of facts that appears beside search results for a recognized entity. But its deeper job is disambiguation: separating your brand from every other thing that shares its name, and fixing which real-world entity a query means.
How a knowledge graph is built
A graph isn’t authored in one place; it’s assembled from evidence that agrees with itself across the web:
- Structured data. Schema.org markup on your pages states your entity and its attributes in a form machines read directly.
- Trusted references.Curated sources such as Wikipedia and Wikidata act as high-confidence anchors for an entity’s core facts.
- Corroborating signals. Consistent mentions of the same facts across independent sites raise confidence; contradictions lower it.
- Relationships. Links to the people, products, and category an entity belongs to give it the context that distinguishes it from its namesakes.
Its role in AI search
The knowledge graph is the connective tissue beneath semantic search: once a query resolves to an entity, the system can match by meaning rather than keyword and reason about related things. For AI answer engines, that entity understanding is what lets a model recognize your brand, place it in the right category, and describe it correctly. Strengthening how clearly you exist as an entity is, in practice, strengthening how well AI can talk about you.
Frequently asked
- What is the Google Knowledge Graph?
- It's Google's database of entities — people, places, brands, and concepts — and the relationships between them. Launched in 2012, it lets Search understand “things, not strings,” and it powers the knowledge panels that appear beside results for a recognized entity.
- How do I get my brand into a knowledge graph?
- There's no submission form. You build the corroborating signals a graph is assembled from: consistent facts about your entity across your own site, structured data that names it explicitly, and presence in trusted references like Wikipedia and Wikidata, reinforced by mentions elsewhere on the web that agree with each other.
- How does a knowledge graph relate to AI answers?
- It gives machines a stable, disambiguated understanding of who you are and how you fit your category. That same entity understanding underpins semantic search and helps AI systems recognize your brand correctly, rather than confusing it with a similarly named one, when they generate an answer.