11
Building Networks

Building Networks. First Decisions What do the nodes represent? What do the edges represent? Know this before doing anything with data!

Embed Size (px)

Citation preview

Page 1: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

Building Networks

Page 2: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

First Decisions

What do the nodes represent?

What do the edges represent?

Know this before doing anything with data!

Page 3: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

Examples

Facebook network What are the nodes? Edges?

What can you see if you were able to visualize this network?

What would network statistics mean in this network?

What patterns might emerge?

Page 4: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

Networks with multiple node types

Bipartite graphs have two node types that do not have connections within the type E.g. no people

connected to one another

Graphs can have multiple node types and not be bipartite

Page 5: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

To Build a Network

Step 1: Define Nodes

What are they?

What are the criteria for being included?

Example: Make a network of characters in your favorite TV show. Nodes are characters, but which ones? Main characters? Extras? People mentioned who never appear? Something else?

Page 6: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

To Build a Network

Step 2: Define Edges

What does an edge represent?

What is the criteria for adding one?

Same example: If two characters know one another, we add an edge. What if they only know of one another? What if they obviously know one another but never

interact on the show?

Page 7: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

Handling Large Networks

Some networks may be too big to analyze.

Filtering or sampling will be important

Page 8: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

Snowball Sampling

When working with a large network, choose a starting node

Get that node, its connections, their connections, and so on until the network is the right size for analysis

Problems: Biased toward the part of the network sampled, may miss other features

Benefits: Easy to do, common

Page 9: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

Random Sampling

Randomly select a certain percentage of nodes and keep all edges between themORRandomly select a certain percentage of edges and keep all nodes that are mentioned.

Problems Edge sampling biased toward high degree nodes Node sampling loses some structural characteristics

Benefits Easy Node sampling keeps some network statistical features

Page 10: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

Egocentric Network Analysis

Instead of looking at the whole network, look at the egocentric networks of some nodes.

A different type of analysis than overall network analysis, but it shows the role of an individual in context.

Page 11: Building Networks. First Decisions  What do the nodes represent?  What do the edges represent?  Know this before doing anything with data!

Exercise

Read the first chapter of Pride and Prejudice

Draw a network of the characters with their relationships.