Week 12: Networks, Text, Maps reprise

Introduction to Data Visualization

W4995.010 Spring 2020

​

​

Fun "map" and bubble chart

00 Quiz

01 Networks: node-link, matrix, enclosure

02 Text: analysis, network, timeseries

03 Maps (reprise)

04 Final Project Announcements

​

01

Networks

​

10min: groups of 4 visualize relationships in Hamilton

Unit: Number of lines in song

Post to following slides

Breakout Room 1

Breakout Room 2

Alexander Hamilton

Eliza Shuyler

George

Washington

Other

Aaron Burr

​

Song 4

Song 1

Song 2

Breakout Room 3: Relationship heat map (opacity = # shared lines)

Breakout Room 4: the breakout room where it happens

Breakout Room 5

Song

Characters

Washington

Hamilton

Burr

Eliza

Other

1

2

3

4

Chord graph encodings:

Chord line: character to song link

Line thickness: number of lines in the song

Color: character name

Breakout Room 6

Alex

George

Eliza

errbody else

Legend

Song 1

Song 2

Song 3

Song 4

Networks vs. Trees

​

​

Network = graph of relationships

between discrete objects

​

​

Tree = network with hierarchical

structure

Munzner

Three Main Ways to Visualize

Munzner

Common Applications

​

• Computer Networks
• Social Networks
• Concept Networks
• Gene Networks

Via Marti Hearst / Jeff Heer

First

Munzner

Tree Layout

​

Via Jeffery Heer

Common Layout: Reingold-Tilford “Tidy” algorithm

​

• Clearly encode depth, no edge crossings
• Ordering & symmetry preserved, compact layout
• ~Linear time (binary and n-ary trees)

Via Jeffery Heer

D3 tidy tree

Via Jeffery Heer

Radial = hierarchical tree in polar coordinates

Via Jeffery Heer

Force Directed Layout

​

• Nodes repel each other
• Edges act as springs
• Friction to ease motion
• Techniques to dampen “jitter”

Via Marti Hearst

Bostock on force layout encoding opportunity

​

But scale?

Via Marti Herst, Jeffery Heer

More Nodes, More Problems…

​

• Tree breadth often grows exponentially
• Even with tidy layout, quickly run out of space

​

Possible Solutions

• Alpha
• Filtering
• Focus+Context
• Scrolling, Panning, Zooming

Via Jeffery Heer

Hairballs… or maybe not?

Left: Apple, Right: Google, Periscopic via Fast Co. Design

Chord Diagram with Hover: Uber Rides

Bostock, Block

A different layout approach

Bloomberg Graphics, 2016

Also a network

NYT How the Virus Got Out

Alternatively...

Munzner

Node-link vs. Adjacency Matrix

Cliques: every node is connected to every other node

​

Biclique: every vertex of the first subset is connected to every vertex of the second subset

​

Cluster: a graph whose connected components are cliques

Adjacency Matrix: Les Misérables

Via Marti Hearst

One more way: Arc Diagram

Heer, J. “Visualization Zoo”

Alternatively...

Munzner

Enclosure/Treemaps: filesystem, Schneiderman ‘91

http://www.cs.umd.edu/hcil/treemap-history/

Enclosure/Treemaps: Map of Market, Wattenberg ‘98

Smartmoney.com

02

Text

​

What does this say?

​

​

Text is not preattentive

​

​

SUBJECT PUNCHED QUICKLY OXIDIZE TCEJBUS DEHCNUP YLKCIUQ EZIDIXO

CERTAIN QUICKLY PUNCHED METHODS NIATREC YLKCIUQ DEHCNUP SDOHTEM

SCIENCE ENGLISH RECORDS COLUMNS ECNEICS HSILGNE SDROCER SNMULOC

GOVERNS PRECISE EXAMPLE MERCURY SNREVOG ESICERP ELPMAXE YRUCREM

CERTAIN QUICKLY PUNCHED METHODS NIATREC YLKCIUQ DEHCNUP SDOHTEM

GOVERNS PRECISE EXAMPLE MERCURY SNREVOG ESICERP ELPMAXE YRUCREM

SCIENCE ENGLISH RECORDS COLUMNS ECNEICS HSILGNE SDROCER SNMULOC

SUBJECT PUNCHED QUICKLY OXIDIZE TCEJBUS DEHCNUP YLKCIUQ EZIDIXO

CERTAIN QUICKLY PUNCHED METHODS NIATREC YLKCIUQ DEHCNUP SDOHTEM

SCIENCE ENGLISH RECORDS COLUMNS ECNEICS HSILGNE SDROCER SNMULOC

​

​

Via Marti Hearst

Tag Clouds

​

Pro: Can help with “gist” and initial query formation.

​

Cons

• Sub-optimal visual encoding (size vs. position)
• Inaccurate size encoding (long words are bigger)
• May not facilitate comparison (unstable layout)
• Term frequency may not be meaningful (“the”)
• Does not show the structure of the text

Via Jeffery Heer

Simple alternatives are often better

​

Via Marti Hearst

Added Context: Parallel Tag Clouds, Collins ‘06

Colins, Viegas and Wattenberg, IBM Research

• Word frequency
• lists of words and their frequencies
• Concordance
• the contexts of a given word or set of words
• N-grams
• common two-, three-, etc.- word phrases
• Entity recognition
• identifying names, places, time periods, etc.
• Keyword extraction
• Term Frequency-Inverse Document Frequency
• Text Classification
• Topics, sentiment, etc. etc.

Text Analysis

Techniques

Via Jeffery Heer

Concordance: Word Tree, Wattenberg et. al. ‘07

​

​

Entity and relationships: Gorg ‘07 Jigsaw

​

​

Topic modeling: Underwood, PMLA journal ‘24-’06

Visualizing NASA Research, 1958–2008

OCR. The Whole Brilliant Enterprise (2004)

Quantifying “cultural impact” via media coverage

OCR. The Whole Brilliant Enterprise (2004)

“Heatmap” (character mentions over time)

Via Marti Hearst

Darwin’s Origin of Species

(Hamilton, by Shirley Wu & Pudding)

​

​

03

Maps, reprise

​

​

Questions?

​

Next Class…

Next class: Ethics, What’s Next (in industry/research)

​

• Readings x 2
• boyd + Crawford, “Six Provocations for Big Data”
• Hullman, What Is Visualization Research?

​

• Guest lecture: Prof. Wu on dataviz research

​

Final Project Showcase

​

• 5/5 Tuesday 6–8pm

Lightning round (zoom) + science fair (in rooms)

Invite all your friends! I will send an invite you can forward.

​

• Day-Of: final project page + 3 min. lightning talk

​

• Later: Documentation due 5/11
• Still Monday 11:59pm
• Format is important, follow instructions.

Past years’

STAGE

ENTRANCE

FINAL PROJECT SHOWCASE

Spring 2019 W4995 Intro to Data Visualization

columbiaviz.github.io

​

Thanks to Center for Data, Media & Society

Brown Center for Media Innovation

Computer Science Department

Designing the Perfect Board Game

The Effects of 911

Citibike and Cab Demand in NYC

Understanding the United States Opioid Crisis

How bad is climate change? Much worse than you think

Evolution of Terrorism

Does College Provide All Students with the Same Economic Opportunities?

Diversity in NYC’s Specialized High Schools

Past years’