The Knowledge Graph refers to the knowledge base utilized by Google to optimize its search engine’s results for semantic-search. The information in the Knowledge Graph comes from a range of sources, anywhere from Wikipedia to the CIA World Factbook. Like answer engines, such as Ask Jeeves, the Knowledge Graph display is meant to answer a user’s query without redirecting them to other sites. It was added to the search engine in 2012, beginning with the United States.
Billions of facts about relationships between different objects contained within the Knowledge Graph’s semantic network and are used to understand the meaning of search queries. Since its original release, the Knowledge Graph has been introduced in seven more languages and has reportedly contributed to a decline in page views of several language versions of Wikipedia.
In august of 2014, Google introduced the Knowledge Vault, which is a compilation of facts from across World Wide Web. As opposed to the Knowledge Graph which exclusively pulls information from trusted sources, the Knowledge Vault uses machine learning to rank the validity of its results from high to low confidence. The idea is that the knowledge base isn’t just about public sources, but is focused on comprehensive breadth and depth.