程序代写代做代考 Recommender Systems

Recommender Systems

Social Network Analysis
Visualization
Robin Burke
DePaul University
Chicago, IL

1

Outline
Assignments
Homework 3
Homework 4
File formats
Visualization: General
Visualization: Networks
Overview
Nodes
Edges
Layout
Break
Lab CDM 801

Homework 3
Questions?

Homework 4
No more step by step code scaffolding
Working with bipartite data
Les Miserables, but
Section-section projection
Filtering
Save as .graphml file
Make visualizations in Gephi
Demonstrate image inclusion syntax

Turn in
More files
HTML
Rmd
R
GraphML
PNG

Proposal milestone
Fill out the form
Due next week
Issues
Sampling / selecting huge data sets
Data gathering practicalities
Questions?

Graph formats
Different from how graphs are stored internally
How are graphs stored in files?
Many different options

Most common
edgelist
very simple, just a list of edges
no place to store node and edge attributes
Pajek (.net)
(what were they thinking!)
lists nodes and edges separately
also can store multiple networks at once
associated with the Pajek tool
GML (.gml)
stores a bit more graphical information
format a bit like JSON
ids must be numeric
GraphML (.graphml)
an XML format
most comprehensive
least space-efficient
GEXF
similar to GraphML
Gephi project

In igraph
g <- read_graph(“my-graph.graphml”, format=“graphml”) write_graph(g, “new-graph.net” format=“pajek”) Supported formats edgelist pajek graphml gml plus dimacs, dot, leda, ncol, lgl not gexf created by the Gephi project