Fact-Based Models and Graph Schemas

Overview

Recall questions on the fact-based model and what one fact holds, nodes, edges and properties, and how a graph schema stores relationships a relational schema would join for.

What you need to know

Answer each one before you reveal it.

Questions from this subtopic

Each one with the model answer underneath.

1. 1

   What is a fact-based model?

   Model answer

   A way of representing data as a collection of individual facts rather than as records in tables. Each fact stands on its own as one small statement about the world, and the dataset is simply the accumulation of all of them. Nothing is folded into a row with a fixed set of columns, so the model can hold whatever turns up without a layout having been agreed in advance.

2. 2

   What does a single fact capture?

   Model answer

   One piece of information, and no more. Rather than a customer record carrying a name, an address and a date of birth together, each of those is held as a separate fact about that customer. Keeping facts to a single piece of information means each can be recorded, stored and examined independently, and a new kind of fact can be added without disturbing anything already stored.

3. 3

   Why does a fact-based model suit data that keeps arriving?

   Model answer

   Because new facts are simply added to the collection. There is no row to locate and update and no layout to extend when an unfamiliar kind of information appears, so a stream can be written down as it comes without pausing to work out where it belongs. Sources producing quite different information can feed the same store, since each contributes facts rather than having to fit a shared table.

4. 4

   What advantage does recording facts have over updating records in place?

   Model answer

   Nothing is lost. Updating a record overwrites what was there, leaving only the current position and destroying the history, whereas adding a new fact leaves the earlier one in place — so the record shows not only what is true now but what was true before and when it changed. That matters for analysis, where the way a value moved is often more informative than the value it happens to hold today.

5. 5

   What is a graph schema used for?

   Model answer

   It captures the structure of a dataset in terms of the things it contains and the connections between them. Rather than describing tables and columns, it describes objects and the relationships joining them, which suits data whose value lies mainly in how items are connected. It gives such a dataset a described structure without forcing it into the rows and columns a relational schema would demand.

6. 6

   What is a node in a graph schema?

   Model answer

   A node represents one thing in the dataset — a person, a place, a product, an event. It is the entity that relationships are drawn between, and it holds the information belonging to that individual thing. In a graph describing a road network the nodes would be the junctions; in one describing a social network they would be the members.

Other ways to work this

The same content, four other ways.

What is inside Levo

Six ways to work the same material.