AI Lexicon — K

Knowledge graph

People say the concept of a knowledge graph was popularized by search engine company, Google, when it launched its Knowledge Graph in 2012. But knowledge graphs already existed in computer science. It's the idea of linking real-world things, known as "entities" — such as people, objects, facts about people and objects, buildings, places, ideas — and describing the relationships between those entities.

A knowledge graph has three main elements: nodes, edges and labels. "Nodes" are the things or entities, "edges" describe the relationships between those things, and "labels" refer to attributes — the factors that define those relationships.

For Google, it's a case of viewing a search query such as "Taj Mahal" as more than just a so-called "string" or words ("Taj" + "Mahal"). Instead, it analyzes how those words are connected and related to one another, and to other things in the real-world. The aim is to deliver search results that offer more than information that is specific to the Taj Mahal, a monument in India, but to include links to other monuments in India or information about the city where the Taj Mahal stands, Agra, or, as Google wrote in a 2012 blog, India restaurants near you that are called "Taj Mahal". (za/fs)

  

Sources:

What is a knowledge graph? (IBM) https://www.ibm.com/topics/knowledge-graph (accessed August 3, 2023)

Introducing the Knowledge Graph: things, not strings (Google) https://blog.google/products/search/introducing-knowledge-graph-things-not/ (accessed August 3, 2023) 

Read the rest of DW's AI Lexicon:

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

We're keen to hear your feedback. Suggest an entry by sending us a comment. And let us know if you feel we have missed something, got it wrong, and tell us whether our AI Lexicon has helped you understand the technology better.

Written and edited by: Zulfikar Abbany (za), Fred Schwaller (fs)