Google Knowledge Graph

A knowledge graph is a knowledge database in which a variety of information is stored. These can be used to establish relationships among different topics. At the same time, they can be linked or enriched with certain attributes or properties.

Design and structure

The basic structure consists of nodes (entities - usually a name or label) and edges (usually properties or attributes). These are related to each other.

Knowledge Graph

Facebook, for example, uses a social graph to analyze relationships among different profiles or users. Google has been using a link graph for a relatively long time, so that relationships among web pages or files can be analyzed and evaluated.

The Knowledge Graph is responsible for the analysis between entities.

Example:

You are searching for Thomas Müller - relations to other information are available through Knowledge Graph (e.g. Bayern Munich player, date of birth, facial appearance, salary, height, etc.).

The basis is formed by different levels, which are supplied by a huge knowledge repository (Knowledge Responsity) from various sources.

Information from Google

Where does Google get the data for the Knowledge Graph?

Google can obtain the information about the entities and their relationships to each other from many different sources.

Sources for unstructured data

  • Websites
  • Search queries
  • Browser (Chrome) users
  • unstructured databases

Semistructured data

Use of encyclopedias (e.g. Wikipedia)

Structured data

Here, Google uses semantic databases such as.

  • YAGO
  • Google my Business
  • CIA World Factbook
  • Wididate
  • DBpedia
  • Structured data via Microdate, RDF and JSON-LD
  • Licensed data
  • etc.

Google receives information via many different paths. However, the multitude of possibilities also means that it is very easy to differentiate which sources are trustworthy. Therefore, it is also important to control corresponding data etc. and to keep them synchronized (as far as possible).

What changes does the use of the Google Knowledge Graph entail?

By using the Knowledge Graph, the results of the searches have changed a lot. In the beginning only "categories" were available - after that there are around keywords.

In the meantime it became a sematic search.

Example:

Let's stay with the upper example of Thomas Müller. If you search for Thomas Müller today, you will see a lot of information assigned to him in the desktop view on the right side. Here are some examples:

  • Brief description
  • Born on and where
  • Size
  • Salary
  • Spouse
  • Current teams
  • Books
  • Statistics
  • Often searched in context

In addition, the current headlines are also displayed as a carousel in the search results. This already provides the searcher with a large number of answers, so that he can quickly and easily find the information he is looking for.

Google Knowledge Graph

Consequences for providers and website operators

These changes have also created new requirements for providers:

  • It is enough not more, only on a Keyword to optimize
  • Not every search will also load on a web page -. many search queries are handled directly by Google answered (will become even more if the number of searches without display (e.g. by Alexa, etc.) continues to increase)
  • Data and information must be prepared in a more conscious and targeted manner
  • You have to use the Contexts understand
  • Good and quality content to develop becomes more costly
  • The Online reputation must be monitored as the trustworthiness of the information source becomes more and more important for Google.

Summary

Information providers such as Google or Facebook can now link information very well with each other thanks to the very large collection of data. Due to the users' demand to get information faster and easier, they have also developed further and found new solutions.

It is becoming increasingly important to build up a good reputation on the Internet and at the same time to offer high-quality information, which should be prepared in such a way that the information providers can also work with it in a meaningful way.

We will be happy to help you!