The purpose of visualization is insight, not pictures

Ben Shneiderman

How to make sense of data?

• Statistical Analysis
• Machine Learning and Artificial Intelligence
• Visual Analytics (and data analytics)

Why should we visualize?

Datasaurus!

In Infovis we look for Insights

• Deep understanding
• Meaningful
• Non obvious
• Actionable
• Based on data

How do I do it?

What do I use?

Insights

What car should I buy?

Normal procedure

Ask friends and family

Problem

That's inferring statistics from a sample n=1

Better approach

Data based decisions

Social Networks

Twitter election analysis

Presidential Election

Influentials?

How can you do it?

You get a new dataset

What do you do with it?

Datos Abiertos

Tableau?

Jupiter Notebook/R?

Voyager2?

Voyager2 with 50MB

Parallel Coordinates?

SPLOM?

Navio

A viz Widget

const nv = navio(d3.select("#navio"), 600);

nv.data(data);

nv.addAllAttribs();

nv.updateCallback( sel => doSmthng(sel) );

Evaluation

• Usability study
• Data Scientists Experiment
• Domain Experts Validation

Usability Study

MoMa Original search interface

Usability Study

Insight based experiments

Spinoff projects

Shipyard

TADAVA

TADAVA Architecture

Nautilus

Can we use Navio with Voyager?

Stand alone Shipyard

Can we use more resources for Shipyard running locally?

Thesis

2 Msc, 3 Undergrads

Free (as in Libre) Software

• Navio
• Shipyard
• TADAVA
• Backviz*
• Biocicle*

Do you ML?

Multivariate Data?-> Dimensionality Reduction + Clustering

MLExplore.js

Opening the black box

Rappi on Twitter

• 30k tweets in the last 7 days

It's up to you!

• Interactivity 👉 Ask questions
• Slice and dice
• Overview first, Zoom/Filter, then details on demand

😡😠😒😐😏😃🥰?

• Machine learning 🎩! ???
• Detects sentiment ! ???

I hired a data 🐒 (might be me)

Analyze 180 tweets

• 😡😠😒😐😏😃🥰

Here are some of them

Would you hire this data 🐒?

Well.... actually

• It wasn't a data 🐒
• It was a 💻
• Would you use it?

More?

Well.... actually

• It wasn't just any computer
• It was service offered by Google
• It costs $$$

Will you trust it?

• Spanish is just too complicated
• These models are trained with 🇪🇸 examples

¡No coma Machine Learning, coma 🍖!

Industry

Take home messages

Focus on insights!!!

We need more open data!

How can I get Insights too?

No need to wait for Stanford, MIT or Berkeley to help you

IMAGINE Research Group

• Visual Analytics
• Virtual/Augmented Reality
• Visual Computing
• Mobile Robotics
• Machine Learning

Bonus

Big Data?

You might have heard of the Vs of Big Data

• Volume
• Velocity
• Variety
• and Veracity and Value

Too ambiguous!! 🤦🏽‍♀️ Let's go beyond that

How Big is big?

Can you fit it in one computer?

Yes? 👉🏼 Then, is not really big 🤷🏽‍♀️

Why this criteria?

Big data 👉🏼 Big overhead

Example: photo collection

• One photo 👉🏼 10MB
• 1k photos in a 📱 👉🏼 10MB * 1k = 10000MB = 10GB
• 50k photos in your 💻 👉🏼 10MB * 50k = 500GB

Big Data? 🙅🏽‍♂️

How many blue photos are in my collection?

How do you compute this?

• Put all your photos in one 💻
• Go through all the collection and count the blue ones

Flickr scale

80+ trillion photos (80'''000''000'000.000)

That's big data

How many blue photos are on Flickr?

How do you compute this?

• Distribute the data among 100s of 💻💻💻s. (a cluster)
• Compute subtotals on each data part. (Map)
• Aggregate the subtotals into one big total. (Reduce)

How many computers do you need?

What if one computer breaks? ☢️

Conclusion

Big Data? 👉🏼 Only if it doesn't fit on one 💻

⚠️ Use it only if you must ⚠️

But don't panic!

Let me share a secret

🤫

My wife tells it to me all the time!

Size doesn't really matter

What matters are the insights 👍

Insights ?

Making Sense of Data

Anti-corruption referendum

What about the oposition?

Other Insights

FDA

Task: Change in drug's adverse effects reports

User: FDA Analysts

State of the art

Health insurance claims

Task: Detect fraud networks

User: Undisclosed Analysts

Clustering

Overview

Ego distance

My Facebook