Module 3: Entity & Knowledge Graph

35 minutes

What Is the Knowledge Graph?

Google's Knowledge Graph is a massive structured database of entities and their relationships. As of 2023, it contains over 500 billion facts about 5 billion entities.

An "entity" in Knowledge Graph terms is any thing that can be uniquely identified: a person, organization, place, concept, product, creative work, or event. Every entity has:

• A unique identifier
• A set of attributes (properties and values)
• A set of relationships to other entities

Example entity record (simplified):

Entity: Bill Slawski
Type: Person
Attributes:
  - Occupation: SEO researcher, author
  - Notable work: SEO by the Sea
  - Affiliation: Go Fish Digital
  - Born: 1952
  - Died: 2022
Relationships:
  - WorksFor: Go Fish Digital
  - Created: seobythesea.com
  - KnownFor: Google patent analysis

How Google Builds Entity Associations

Google associates web pages with entities through multiple signals:

Structured data signals: Schema markup explicitly declares entity relationships (Person, Organization, LocalBusiness, etc.)

Named entity recognition: NLP systems extract entity mentions from plain text — without any markup required

Co-citation patterns: Pages that appear alongside the same entities in many contexts get associated with those entities

Link anchor text: Links pointing to a page with "Bill Slawski SEO" as anchor text associate the destination page with the Bill Slawski entity and the SEO topic

User behavior patterns: Search queries that mention an entity, then click to a page, associate the page with the entity in the query

Entity Types for SEO

Understanding entity type classification helps you optimize the right signals:

Organization entities: Your business. Key attributes Google tracks: official name, founding date, founders, headquarters, industry, products/services, key people.

Person entities: Authors, executives, founders. Key attributes: full name, profession, notable works, organizations associated with, education.

Place entities: Physical locations, service areas, named landmarks. Key attributes: coordinates, address, contained-in relationships.

Concept entities: Abstract ideas (PageRank, E-E-A-T, semantic search). These are topical entities — your content can become associated with them.

Product entities: Software, physical products, services. Key attributes: manufacturer, category, price range, features.

Disambiguation: The Critical Challenge

Many entity names are shared by multiple entities. Google must determine which entity is intended from context.

High-risk disambiguation challenges:

• "Mercury" — element (Hg), planet, car brand, Roman god, music band
• "Java" — programming language, island in Indonesia, coffee
• "Python" — programming language, snake species
• "Slack" — messaging platform, adjective meaning loose

How Google disambiguates:

1. Context words surrounding the entity mention
2. The page's overall topic (classified via phrase co-occurrence)
3. Co-occurring entities (Python appearing near "Flask" and "Django" = programming language)
4. The site's general topic category (a tech blog mentioning "Apple" = likely the company)

Your responsibility: Ensure every polysemous entity in your content is anchored to the correct meaning through context. This is operationalized in the Entity Extraction Audit.

Why Knowledge Graph Association Matters

Pages associated with strong, established entities inherit some of that entity's authority for related queries. This is why:

• Having your author entity in the Knowledge Graph helps all their articles rank
• Being co-cited with established entities in your niche helps Google understand your topical authority
• Getting mentions on Wikipedia/Wikidata (primary KG sources) strengthens your entity's presence

The practical upshot: entity optimization is not separate from content optimization. It's what happens when you consistently write about a topic with enough depth and consistency that Google's systems confidently associate your site/author/brand with that entity.