Graphs

Our Earth Digital Twin will be modeled on a knowledge graph. Knowledge graphs are a collection of interlinked descriptions of entities and objects. While the standalone objects are important themselves, it is from understanding the relationships between those objects that provide the most useful insights which makes the knowledge graph so powerful.

Google launched its version of a knowledge graph in May 2012 which allows Google to improve the relevance and accuracy of its search results greatly. It goes beyond simply matching keywords to queries and instead delivers results that better understand real-world entities and their relationships to one another.

Facebook also uses a knowledge graph to improve its search function by connecting relationships among its users through the information it collects about objects such as the music, movies, celebrities and places that interest its users. 

Netflix uses knowledge graphs to arrange information on its huge catalog of content, inferring links between TV shows, movies and directors and actors. The knowledge graph then helps infer what users might like to watch next, and promote the binge watch business model.

Siemens uses knowledge graphs to construct models of the data it produces and stores and employ it for risk management and process monitoring applications. They also use knowledge graphs to build digital twins which is a simulated form of real world systems and use the graph to design, prototype and train. Knowledge graphs are also being used in financial sectors for monitoring fraudulent transactions and for tasks such as investment analytics and marketing.

Due to their rapidly growing scale, the storage and maintenance of most real world knowledge graphs is becoming challenging. By May 2020, the Google knowledge graph had over 500 billion facts on 5 billion entities.

To deal with large scale knowledge graphs, the AI community has been using machine learning, not only to quickly build and structure knowledge graphs, but also to infer links between data points that would not be noticed otherwise. However, machine learning is also getting benefits from knowledge graphs to more deeply understand data such as text, video and audio that cannot be fitted well into the relational database.