week 4

Question 1

network

Answer 1

relations/connections can be drawn

used for any complex system of interrelated things

Question 2

things are called ___

connections are called ___

Answer 2

nodes

edges

Question 3

nodes and edges

Answer 3

can be anything

nodes are people and edges are friendships or followers

Question 4

when are networks useful

Answer 4

reduce or understand complexity: we can reduce complicated data to its overall structure

allow us to spot patterns and make inferences on the structure of our data

allow us to consider how data is interrelated and how relationships between nodes effect each other and the overall structure

Question 4

edge weights

Answer 4

connections can often have a weight attached,

ie, number of letters exchanged between two people or number of times two actors appear in a scene

Question 4

edge direction

Answer 4

directions, meaning that the incoming and outgoing links are counted seperate

ie, facebook vs twitter frienfs network

Question 4

node level metrics

Answer 4

degree

• count of a nodes connections
• most basic measurement of importance in a network (known as centrality)

Question 4

can be directed

Answer 4

Question 4

can be weighted

Answer 4

Question 5

or both

Answer 5

Question 6

betweenness centrality

Answer 6

It measures how important a point (node) is in a network.

A node has high betweenness if lots of paths go through it.

It acts like a bridge or connector between other nodes.

ie, think of a network like a road map:

If one city is the main stop between many others, it has high betweenness.

If that city is removed, it disrupts travel for many routes.

Question 6

why visualise a network

Answer 6

descriptive data analysis, as a way of describing the overall structure of graph

exploratory data analysis, when its used as a sort of map to understand its various components and help spot patterns or interesting features by eye

Question 7

force directed network visualisations

Answer 7

These graphs use simulated physical forces to place the points (nodes)

Question 8

force directed network advantages

Answer 8

Helps you see structure in complex data.

Shows clusters, key connections, and outliers.

Useful for exploring data when you don’t yet know what patterns to expect

Question 8

How to Read a Network Graph

Answer 8

Look for clusters (groups of nodes close together).

Notice empty areas (sparser parts of the network).

What’s in the centre? What’s on the edges?

Check how colour, size, and shape are used — they might show different categories or values.

Axes (x and y) usually don’t have set meanings — it’s about relationships between points.