What Is Google’s Knowledge Graph, and What Is It Used For?

Google’s Knowledge Graph is a structured knowledge database that organizes “entities,” such as people, companies, and places, along with their relationships. It enables the search engine to go beyond matching keywords to understand what a query refers to and how things are related, supporting semantic search and the presentation of relevant facts. Knowledge panels in search results are one interface for displaying this information, not the Knowledge Graph itself.

Connecting names, attributes, and relationships into meaningful information

To understand the Knowledge Graph, think of it as a network of information with clearly defined connections rather than a list of keywords. This network contains three basic elements:

• Entities: Identifiable things, such as Apple, Steve Jobs, and the iPhone, each represented by a separate node.
• Relationships: Connections between nodes, such as the “founded” relationship between Steve Jobs and Apple, or the manufacturing relationship between the iPhone and Apple.
• Attributes: Specific details about an entity, such as a company’s official name, founding date, and headquarters location.

Semrush explains that each entity has a unique identifier, which helps distinguish things with the same name. This means Apple the company and apple the fruit need not be treated as the same thing despite their similar names. The focus is on identifying what a name refers to, not merely whether the words match.

Searching without knowing the exact name of what you are looking for

One use of the Knowledge Graph is to help Google understand the underlying meaning of a query. Semrush gives the example of a user asking, “When was the company that makes the iPhone founded?” Even without Apple being named directly, Google can understand that the query concerns that company’s founding date.

This is a form of semantic search: using the relationship between a product and a company to identify the entity a question refers to, then finding the required attribute. However, the Knowledge Graph is only one of the methods Google uses to understand queries; it should not be credited with all of Google’s search understanding capabilities.

Knowledge panels show summaries, with data drawn from multiple sources

The Knowledge Graph is the backend database, while a knowledge panel is the visible information card in search results. For a company, a panel might list a brief description, its founders, and its founding year, helping readers grasp basic facts without having to switch between multiple webpages to find scattered information.

According to Semrush’s overview, Google obtains information from public sources, licensed data, and content owners. Public sources include Wikipedia. When the system determines that a query calls for information about a specific entity, it may display that information directly in search results. Knowledge panels may also incorporate content from other sources.

When reading a panel, it is therefore important to distinguish between the system that organizes information and the interface that displays it. A panel provides a summary of an entity, not a complete account from any single website, and its content should not automatically be assumed to have been written by the person or organization it describes.