What is a graph and why is it a powerful way to represent knowledge?

"In many ways, the most elegant, simple, and powerful way of representing knowledge is by means of a graph. A graph is composed of a bunch of little bits of data, each of which may (or may not) be attached to each of the others. Each of the labeled ovals is called a vertex (plural: vertices), and the lines between them are called edges. Each vertex does, or does not, contain an edge connecting it to each other vertex. One could imagine each of the vertices containing various descriptive attributes — perhaps the John Wilkes Booth oval would have information about Booth’s birthdate, and Washington, DC information about its longitude, latitude, and population — but these are typically not shown on the diagram. All that really matters, graph-wise, is what vertices it contains, and which ones are joined to which others.\n\nCognitive psychologists, who study the internal mental processes of the mind, have long identified this sort of structure as the principal way that people mentally store and work with information. After all, if you step back a moment and ask “what is the ‘stuff’ that’s in my memory?” a reasonable answer is “well I know about a bunch of things, and the properties of those things, and the relationships between those things.” If the “things” are vertices, and the “properties” are attributes of those vertices, and the “relationships” are the edges, we have precisely the structure of a graph. Psychologists have given this another name: a semantic network. It is thought that the myriad of concepts you have committed to memory — Abraham Lincoln, and bar of soap, and my fall schedule, and perhaps millions of others — are all associated in your mind in a vast semantic network that links the related concepts together. When your mind recalls information, or deduces facts, or even drifts randomly in idle moments, it’s essentially traversing this graph along the various edges that exist between vertices. That’s deep. But you don’t have to go near that deep to see the appearance of graph structures all throughout computer science. What’s MapQuest, if not a giant graph where the vertices are travelable locations and the edges are routes between them? What’s Facebook, if not a giant graph where the vertices are people and the edges are friendships? What’s the World Wide Web, if not a giant graph where the vertices are pages and the edges are hyperlinks? What’s the Internet, if not a giant graph where the vertices are computers or routers and the edges are communication links between them? This simple scheme of linked vertices is powerful enough to accommodate a whole host of applications, which is why it’s worth studying."