Good afternoon

Good afternoon!

Hello!

Hello, good afternoon.

Hi

Good afternoon!

Good afternoon all from Ann Arbor, Michigan — the smartest place in the US!

Hi. There is no sound?

Hello everyone!

Hello all!

Good afternoon!

Hi everyone

Good afternoon. The audio works well

Good afternoon

Good afternoon from Santiago, Chile

I have sound

Tina: I have sound

Hi everyone; sound is fine for me

Good afternoon

I can barely hear the speaker

Hello

Found it!

I hear the speaker well

Hi everyone.

Good morning

The audio is ok 👍

Hello

Hey guys

Hello everybody. Martha Garcia from Honduras

Good morning from Utah

Good morning! Sound is good. Jeff Heer is doing the intro now and is clear.

Good Morning.

Audio is OK

Hello from NJ

Good afternoon from Portugal.

hi there

Hello from Brisbane, Australia

Hi everyone

hello from Kansas

Goof afternoon from Milton Keynes in the UK

Here the audio is also OK.

Good Morning from Arizona!

Hello everyone

Good afternoon

Audio fine. Dang zoom - forgot it kept its own audio output selection.

Hello from Portland, OR

Hi everyone

Hello from Ann Arbor Michigan

Hello from Zagreb, Croatia!

hi, Thomas from Vienna in Austria

cool - happy to see a recording happening...

Heloo everyone. This is Gustavo from Recife, Brazil.

Hello from Ann Arbor

Hello from Washington, DC

Hello from Richard in Oakland, CA.

Good evening, its 5pm from Ghana.

Hello from Minneapolis

Stanford and UBC -- what a great set of speakers

Hi from Raleigh, NC

Hello from NJ

Hello from Palermo, Italy

Hello from Washington, DC

Hello from Saskatoon, Canada!

Hi from Madrid, Spain

Hi from Switzerland 🙂

Hello from Los Angeles

Hello From Rome, Italy

Hi everyone, from Belfast UK.

Vancouver, Canada

Hi from Tempe, Arizona!

Hello from Omaha, NE

Hi from Barcelona

Hi, will the talk be available offline also?

Hello from Hochiminh, Vietnam

Hi from Switzerland

Hello from Campinas, Brazil.

Hi from Guatemala

Hi from Giessen, Germany

Hi from Barcelona!

Greetings from Seattle!

Hello from Berkeley, CA!

Greetings from Lubbock, TX!

Hi from Vancouver, Canada

Hi, Dresden germany

Hello from Ghana

Hi from Athens Greece

Hi from Aachen, Germany

Hi from Seattle

Hello from Los Angeles, CA

Hi from Duluth MN, USA

Hello from Bangalore, India

Hi from Sarasota, FL

Hi from Knoxville, TN (US)

Hello From Boston, MA

Hallo from Munich, Germany 😊

Hello from AZ.

Hi from Aachen, Germany

Hi from Paris

Hello from Hong Kong!

Hello from Golden, CO!

Hi from India

Hello from Rio de Janeiro

Greetings from Lubbock, TX!

Hi from Montréal!

Hello from Kingston, Canada

Howdy from Texas!

Hello from Knoxville, TN (USA)

Hi from Oak Ridge, Tennessee!

Hi from Doha, Qatar!

Howdy from Texas!

Greetings to all from Ioannina, Greece

Hi from Rennes, France :-)

Joining from Boston, Mass - hello everyone!

Hello from Travelers Rest, SC!

Hello from Kingston, Canada

Hi from Bogotá, Colombia. :-)

Hi Everyone, from the desert in Tempe AZ

it's perception that matters so visual activity important.

Hello everyone from Colorado, USA.

Hello from Acton, Massachusetts

Hi from Munich in Bavaria!

Hi!

Hi from Massachusetts, USA!

Aloha from Hawaii

Hello from Lester Cowley in Cape Town, South Africa

Huntsville, Texas, Hello to everyone!

hello from phx

Hi all!!

No slides? For a talk on visualization?

Good afternoon from Yorktown Heights, NY.

Hello everybody what a wonderful day

Hi from Tamaulipas, México!

Hello from Buenos Aires!

Hi All !!!

Hi from New Hampshire, USA

Hello everyone

Hi, from Manipay, Jaffna. Srilanka

hello everyone

Greetings from Chicago

Hi from Munich 💛💙

Hi from São Paulo, Brazil

Greetings from the Netherlands

Welcome, everyone! If you would like to submit a question for the Q&A today, please use the Q&A button rather than chat. Thanks!

Hi from São Paulo Brazil

😃

Good morning from SF Bay Area

tech expert https://www.linkedin.com/in/jimmyrcom/

Hello everyone! From Huntsville, texas.

Hello from India

Hello everyone, from London, UK

Hello from India

Hello from Thessaloniki, Greece!

And I started with a Chemistry major....

Hello from Munich, Germany

Hello everyone, from London, UK

Hi from USA

Hello from the planet Mars.

hello from moon!

Hello from Tallinn, Estonia!

hello from Chile

Checking in from Minneapolis, Minnesota

See a lot hello, so also from my side: hello from Cracow Poland

Hello from Atlanta, GA

Port Washington NY #cybersecurity viz

Hello from Vancouver, Salish Sea, Canada

Hi somewhere next to Paris, France

Hello from Tillamook, OR, USA

Hello from Vancouver, Salish Sea, Canada

Hi! From Salamanca (Spain)

Howdy from North Houston, TX metro.

Hi Everyone, I'm from Vancouver, WA

hello from Cambridge uk

Hello Everyone! from Johannesburg, South Africa

Hello from the Wirral, UK: data analysis and visualisation for wireless telecoms worldwide

Hello from Mt. Kisco, NY (about an hour by train north of Manhattan. I just finished teaching a class.

Hello from Victoria BC

Hello from Phoenix, AZ <3

That's funny, I'm a grad student at UMN. Math PhD.

https://en.wikipedia.org/wiki/Pl%C3%BCcker_coordinates

Hello from Victoria BC

Hello From Madrid, Spain. Hi, Paul! Long time no read! 😅

long time no read!😂

Hello from Houston, TX, USA.

Hello from the pinky of the mitten state.

Hi--could you please paste your question in the Q&A? Thanks!

I’m a UMN alumna

Thanks!

Hi from Calgary AB. Loving this discussion. Inspiring.

Enjoying this fireside chat from Pune in India :-)

Welcome! Glad you're enjoying.

Pat Hanharan referenced this talk a couple of mins ago : https://www.youtube.com/watch?v=SKVI-3K8HNw

Thanks Murali

^ Thank you!

cant agree more!

Thanks Murali

Thanks again Murali

What was the name of that book?

Thank you, Murali

Data Feminism by Catherine D'Ignazio and Lauren Klein

Data Feminism @Michelle Colder Carras

Thanks!

about data, see : https://www.romeolupascu.net/2020/08/23/real-fact/

So visualization is making something visual. Question: can you learn more by watching data stream past you - seeing one point at a time - or 'see all the data together as a table'? I think you see anamolie better when you see the data in a table. 'jumps' in process stand out. You can see right away what is big and what is small. With streaming data, it would take memory and take longer to recognize distinctive points.

The real benefit of visualizations is using the lines, boxes, dots, connections, directions that are in space, on a page or a screen, to think. Lines show relationships, boxes show inclusion, sets—far more accessible and direct than words or numbers

We just published results of our 3-year NSF-funded research comparing predictive accuracy between human-centric system and machine model: http://ssrn.com/abstract=3981732.

https://mitpress.mit.edu/books/data-feminism

Nowadays, we got PB class DBs commonly used. The question is how to summarize the data, n graphs with effective zoom in zoom out, while summarizing the vertices and edges, and avoid a blizzard. Bring in temporal walk, making a movie of progression in time. We need to have rich semantics. For example, connecting a domain specific ontology to the data graph, and look at it from semantics viewpont.We need AI help to achieve this. Still today, we see our tools are primitive for this task

Q: To what extent is data analysis “intuition” vs “insight”?

Hi Romeo

Convincing people of a particular analysis is where I find data visualization especially helpful/useful

Hi Pat, Thanks for teaching Computer Graphics at UW. I remember the hardware we used was over in Meteorology and Space Science.

Humans won handily vs ML in small data technically complex setting, even using small crowds of laypeople

Hi Barbara!

Hi Barbara!

Regina — fascinating!

and yet we still can't plan risk effectively without using milimorts.

Monte Carlo was actually what my non-scientist friend recommended

Hello from New York City

Hello from Atlanta, GA :)

In 1980 at The University of Kansas we held a course named, “Scientific Modeling and Data Visualization” sponsored by a geography professor and two design professors. What took you so long?

I wonder how much of this is cultural. Are there other cultures who are better or more comfortable with uncertainty? Could we do something in our schools to help people be more confortable instead of having a black/white view of the world?

Newer Hawkeye is now showing the image of the ball hitting the court.

I vividly recall a comment Prof. David Draper made, "statistics is the quantification of uncertainty." https://engineering.ucsc.edu/people/draper

I totally agree with your point about the important of causality analytics over the usual predictive/descriptive analytics (pattern matching). What will be your recommendation to represent the stimulant-cause relationship in a generic causal discovery analysis

partially

Dang, maybe I do have to go get another degree then…

Good point Maxim

😃 Question to professors: Human as a 3 dimensional creature, only understands 1D, 2D, 3D visuals best. Do you think the best practice / solutions using visualization is (or almost) thoroughly explored? What are the areas yet to be explored in visualization?

hi Joe, :)

ethnical measure of model is depending on the use/purpose of the model

Per Pat's comment, the commonly used tracks in hurricane forecasts actually does not show the probability of the track nor intensity from an ensemble of models that does have information about the probability distribution. That visualization shows a range derived from past storms. That's why the track is symmetric with respect to the track and a width that expands monotonically with lead time.

@Yaying Zhang,Use the Q&A tab at the bottom of zoom for questions. I think they're not taking questions from chat.

some take on the label 'AI" https://www.romeolupascu.net/2018/05/03/ai-or-aa/

“All the models are wrong, but some are useful” G.Box

agree!

Biggest problem in using economics for policy - unintended consequences. Ubiquitous.

understand the data before any building of the model

the use and abuse of models given the surge of interest in the field it was bound to happen

Great point, Tamara!

NOAA has an excellent visual simulation of sea level rise with which you can explore the effects of low, medium, and high end predictions on specific coastal areas. This enables non-expert users to see what could happen in areas where they live or travel. It can help inspire confidence in what the science is saying.

great 👍

Saw Fanny Chevalier mention this misleading chart in a new YouTube video - https://www.livescience.com/45083-misleading-gun-death-chart.html

Is there any work being done in automated visual analysis along the lines of visualizations fact checking?

👍 true

Something that would immediately flag these kinds of charts being propagated on social media

Interesting example of hand-drawn infographics: https://qz.com/906774/w-e-b-du-bois-commissioned-beautiful-hand-drawn-data-visualizations-and-infographics-for-the-paris-world-fair/

the "funniest" thing ever in data processing is the notion of "data cleaning", a term too darn close to "data laundering" ... and ignoring the problem of our inability to control the error levels in our data, (another form of GIGO) https://www.romeolupascu.net/2017/12/31/fact-fiction-and-the-truth/

or maybe come up with a kind of normalized representation that allows one to compare apples to apples wrt different kinds of visualizations and charts

Great points!

interesting point, RL. It may occur when the researcher is still learning

validation

From an anthropologist and practical ethnographer and data geek — thanks!!!

I personally enjoyed using Julia (https://julialang.org/) for interactive data visualization recently

Data “cleaning” is even more basic… Canadian post codes like A1B 0C1 are often miskeyed or improperly entered. If a pattern (as in Canadian post codes) can be identified, “dirty” datapoints can be spotted. If no pattern or rule, it is much harder.

https://monachalabi.com/#

I uses Mathematica to analyze and visualize data and build circuits that monitor and control devices. LinkedIn with me at https://www.linkedin.com/in/pbrane/ and let me know how you visualize data.

Charles Joseph Minard's Napolean's Russian Campaign comes to mind as a great hand-drawn visualization. 1869.

Michelle, there’s an interesting article on uncertainty and trust here:

https://www.pnas.org/doi/full/10.1073/pnas.1913678117

😀

Thanks, @Francisco!

How about text data ?how do we visualize them when its non english .

Would be good to hear more about principles and less about specific tools.

Coming from GIS this is really interesting discussion

tech expert https://www.linkedin.com/in/jimmyrcom/

Great answers.

thanks

I’m visualizing your cat

haha

KITTY!

(data cleaning) how do you know what the information "source" intended?all data "cleaning" are based on some form of "guessing" we call that with funny names like "iner-extrapolation" but in reality is god darn guessing. And all guessed data introduce uncertainty in the final "product" uncertainty no one can tell what is it. Not scarred enough yet?

A cat is more difficult to visualize than, say, a teapot.

cat visualization

data analysis can help with creating cat-e-gories

See https://www.illc.uva.nl/cms/Research/Publications/Dissertations/DS-2002-03.text.pdf foe a detaled analysis of the Minard chart.

Thanks, Peter!

Good idea but does it encompass all the graphs Algorithms?

one of the worse scenario Hollywood made id "normal" is "inventing" data in pictures of low resolution to make them "understandable" by "inventing data!" the so well known word "enhance". And people love it... Then you feed all that in a model and all hel break loose .... Not scared enough yet? https://www.romeolupascu.net/2016/10/23/fake-fiction-and-bs/

Data cleaning example in nature: Cells convert DNA into RNA, and then RNA into proteins.

Jeff, thanks for mentioning me.

Is that the Semiotics of Graphics author?

And Bertin

*Semiology

Kids get engaged in visual analytics very early in their (rather sophisticated!) computer games. There should be more interaction between the vis community and the game developers to explore some of these questions!

making visualization "to please" the recipient is dangerous if by doing so we obscure the big picture and enhance the target audience biases

@roman Lupascu not all data cleaning is guesswork. For example, we can be certain that in age data records for people, entries containing negative ages are wrong. However, you're right in that many other data may be less easy to be certain about.

so how are you "certain"?

I greet you from the Netherlands Roermond

what gives you that guarantee?

My email is eduard.jacob1@gmail.com

One pitfall is that people focus on a single view - analysis is the journey through many views

Domain knowledge is important for effective data cleaning.

And importance of attention to what you get... The gorilla on the basketball court.

Question: I have many colleagues from non-visualization ateas who approach my research group with visualization needs. Most times their needs can be satisfied with established methods, like parallel coordinates

Domain knowledge plays a key role in data cleaning

truth is there is nothing "certain" out there, we always have to deal with uncertainty and yet how many data points you've seen with uncertainty levels attached to them? (I never seen one)

Banish ROYGBIV from all visualizations…

... how do you balance helping these colleagues with doing research?

Recall Lloyd Treinish & Pravda color design tool from IBM DX

@Sireesha, yes! Domain knowledge is key. What looks suspicious in the data?

Understanding what the data actually represents is important. We can get the needful inputs from SMEs. And again at the end of the day there is no “objective” truth over there we are aiming to reach. We try to reduce as much noise as possible to get the signal we are interested in

How do you make business users care/ask for about "better" visualizations - especially beyond visualization just for reporting purposes?

a "bad typed" data is still data, right, just deleting (ignoring it) does not make data "cleaner" it only changes the type of uncertainty you are dealing with so that process is a misnomer when is called "cleaning". But if we convey the idea of "clean" to people that do not understand but use data (a cop for example) ugly things can happen

What’s the most satisfying/delightful visualization in your respective careers?

I have the same question Brandon - copying into the Q & A section now

Thank you so much for a fascinating chat, I could listen to this all day.

it is what we want to present

to give a value and real information

thanks for this wonderful conversation!

This has been seriously interesting — very thought-provoking!

Hans Rohling was very skilled at connecting data to the world

1. Ontology. 2. Data. Not the other way around.

Romeo Lupascu: 'truth is there is nothing "certain" out there' - Are you certain about that?

*Hans Rossling ;)

Mark: totally agree!

*Hans Rossling ;)

@jock mackinlay: for sure! related, I had the good fortune to work with some ex-gapminder folks while working on public statistics at google

Thank you!

👏

thankss

Wonderful presentation! Thanks!

Thank you so much for you

Thanks!

http://[::]:8000/

Great stuff! Thanks!

Thank you!

Thanks

Thanks! Wonderful talk!

Thank you

Thanks!

Thank you!

Thank you! Loved this

Thanks

Thank you!!

Thank you!

thanks

Thank you!

Thanks. Later!

Thank you so much for this!

Thank you

Thanks! Have a nice day.

thanks all, good chat, please share it

thank you

Thank you ALL!!!

Thank you!

Thank you!

Thank you

Thanks!

will there be a online readable transcript?

thanks

Thank you!!

Thanks.

Thank you!

Many thanks!

Thanks - great talk!

Thank you!

Thank you.

Thank you. I really enjoyed the discussion today.

thanks!

Have agood day to everyone from Canada

Thank you

Thanks

Thanks :)

thanks!

Great ideas, thank you!

how many people joined?

Thank you

Thank you!

Thank you! EXCELLENT!

Thank you!

awesome talk, thanks all ;-)

Thank you

Thank you all from Turkey

This was a cool one. thanks

Bye bye!

Thank you

Thanks!

Thank you!

Thank you!

thanks

Thank you, great session.

Thank you!

Thank you!

Thanks! Hi, Tamara, from Asilomar.

👏

Thanks!

Thanks!

Thank you!

many thanks!

thanks!!

Loved this fireside chat! Thank you! :-)

Thank you all!

Thank you!

Thank you 👏

Thank you!

Thank you!

Many thanks!

Thanks

Thank you!

Thank you!

Thank you!

Thank you!

Many thanks

Thank you so much! that was wonderful

thanks

thanks

Thank you

thanks

TY