ThisCalculator

Everyday

Google Knowledge Graph 13 Years Ago?

Updated · Standard formula, shown below

Thirteen years ago, in May 2012, Google launched the Knowledge Graph, a system that turned search results from a list of links into a panel of facts about people, places, and things. It was the first major step toward the “answer engine” we use today, and it introduced the concept of entities and their relationships to search. This article explains what the Knowledge Graph was, why it mattered, and how it still shapes search results.

Before the Knowledge Graph: The Old Search World

Before 2012, Google Search was essentially a keyword matcher. Type “Albert Einstein” and you got a list of blue links ranked by PageRank and other signals. You had to click through to a website to learn anything. There was no sidebar with his birth date, occupation, or spouse. Search engines did not understand that “Einstein” was a person, that he was born in Germany, or that he was related to the theory of relativity. They just matched the string “Albert Einstein” against web pages.

This worked for many queries, but it failed for questions like “How tall is Mount Everest?” or “What is the capital of France?” You had to open a page and scan for the answer. The Knowledge Graph was designed to answer those questions directly, by building a database of real-world entities and the connections between them.

What the Knowledge Graph Actually Is

The Knowledge Graph is not a single database, but a system that aggregates structured data from many sources. At its core, it is a graph database: a network of nodes (entities) and edges (relationships). Each entity has a unique identifier, and relationships are typed. For example, the entity “Albert Einstein” has a relationship “born in” to the entity “Ulm”, and “spouse” to “Mileva Marić”.

Google sources this data from a variety of places, including:

  • Public structured sources like Wikipedia, Wikidata, and the CIA World Factbook.
  • Licensed data from commercial providers like Thomson Reuters and the World Bank.
  • User contributions through the now-defunct Google Contributor program, which allowed users to suggest edits to Knowledge Graph entries.
  • Web crawling of schema.org markup and other structured data on websites.

The system stores facts as triples: subject, predicate, object. For example: (Albert Einstein, born in, Ulm). This allows Google to answer questions like “What is the capital of France?” by finding the entity “France” and following the “capital” relationship to “Paris”.

How the Knowledge Graph Changed Search Results

The most visible change was the Knowledge Panel: the box on the right side of the search results (or top on mobile) that shows a summary of the entity. For a person, it shows a photo, birth date, occupation, spouse, and a short description. For a place, it shows a map, population, and nearby attractions. For a movie, it shows the cast, director, and runtime.

But the Knowledge Graph also affected search in less obvious ways:

  • Disambiguation: Searching “Jaguar” now asks whether you mean the animal, the car brand, or the operating system. The Knowledge Graph groups results by entity.
  • Direct answers: Simple factual questions get an answer box at the top of the results, often without any link.
  • Related entities: The panel often includes “People also search for” and “Related searches” based on graph connections.
  • Voice search: Google Assistant and Google Home rely heavily on the Knowledge Graph to answer spoken questions.

How the Knowledge Graph Works Under the Hood

When you type a query, Google first tries to identify the entity you mean. This is called entity resolution. It looks at the words, the context, and your search history to decide which entity matches. For example, “apple” could be the fruit or the company. If you also searched “iPhone”, the system likely picks the company.

Once the entity is identified, the system retrieves all its properties and relationships from the graph. It then ranks which facts to show in the panel, based on relevance and popularity. The panel is generated dynamically, so it can change as new facts are added.

The graph is constantly updated. Google uses automated crawlers and machine learning to extract new facts from the web. However, not all facts are verified manually; some errors can slip in. For example, a false birth date might appear if a website with incorrect data is crawled.

Common Misconceptions About the Knowledge Graph

A frequent misconception is that the Knowledge Graph is the same as the search index. In reality, the search index is a database of web pages, while the Knowledge Graph is a database of entities and facts. They are separate systems that work together. The index provides links; the graph provides answers.

Another misconception is that the Knowledge Graph is a single, unified database. It is actually a collection of sub-graphs, each covering a different domain (people, places, movies, etc.). Google does not publish the full graph, so we do not know its exact size. Estimates range from billions to trillions of facts, but these are guesses.

People also assume that the Knowledge Graph is always accurate. It is not. Errors occur, and Google has a feedback mechanism for users to report them. However, the process is not transparent, and many errors go uncorrected for years.

The Knowledge Graph’s Impact on SEO and Websites

For website owners, the Knowledge Graph changed the game. Previously, ranking high in the blue links was the only goal. Now, being the source of a Knowledge Panel fact is also valuable, because the panel appears above the first organic result on many queries.

To increase the chance of appearing in the Knowledge Graph, websites can use structured data markup (schema.org) to explicitly tell Google what their content is about. For example, a recipe page can use the Recipe schema to provide cooking time and ingredients, which may appear in a rich result.

However, the Knowledge Graph also reduced click-through rates for some sites. When Google shows a direct answer, users do not need to click through to a website. This is known as the “zero-click search” phenomenon, and it has been a concern for publishers since the Knowledge Graph’s launch.

Frequently Asked Questions

Is the Knowledge Graph the same as Google’s search index?

No. The search index is a database of web pages, while the Knowledge Graph is a database of entities and their relationships. The index is used to retrieve pages; the graph is used to retrieve facts.

How can I get my business into the Knowledge Graph?

You can’t directly submit to the Knowledge Graph. However, you can increase your chances by having a well-structured website with schema.org markup, a Wikipedia page, and consistent information across the web. Google decides which entities get panels based on its own algorithms.

Why does the Knowledge Graph sometimes show wrong information?

Because the graph is built from many sources, some of which may be outdated or incorrect. Google uses automated processes to extract facts, and errors can occur. You can report a problem via the “Feedback” link at the bottom of the Knowledge Panel.

Does the Knowledge Graph affect voice search?

Yes. When you ask Google Assistant a factual question, it often uses the Knowledge Graph to generate the answer. For example, “Who is the president of the United States?” relies on the graph’s data.

Can I see the full Knowledge Graph?

No. Google does not publish the entire graph. However, you can see a subset of it through the Google Knowledge Graph Search API, which was available until 2019. Since then, there is no public API.

What was the first entity in the Knowledge Graph?

There is no public record of the first entity. The system was launched with a large set of entities from Wikipedia and other sources, but Google never revealed the exact list or the first one.

Related

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.