She was on the, what ended up being the second-place team, but in my mind, I think,

was the winning team of the presentations.

Namrata Zalewski, and her presentation is entitled, "Tableau,

Using Visualizations to Truly See Your Data".

It's really a transformative tool.

If you have not used Tableau yet, I thoroughly encourage you to do so.

Also check out ds.iu.edu.

So please give a warm welcome to Namrata Zalewski.

[ Applause ]

>> [Laughing] I'm Namrata Zalewski [applause].

And I will talk about using Tableau for data visualization at Indiana University.

A few weeks ago, some of you may remember that I sent out a survey asking Tableau users

at Indiana University what they think of Tableau.

And here's what I collected.

From 25 participants, 99 words and 4 words per participants.

So what I've done with this data, of course, built a Tableau visualization.

This is a snap of a Tableau dashboards on this slide.

And what you're seeing here are the words that were mentioned once.

And I have highlighted some of the words that I thought were very interesting.

So basically the users are saying that Tableau is intuitive.

It is engaging.

It is challenging.

It allows you to explore data.

It is expensive, and it allows you to do analytics.

The words that were mentioned twice basically says that Tableau is fast.

It has visualization capabilities.

Allows you to see big pictures.

Has drill-down capabilities.

It is flexible.

And it is colorful.

The words that appear three times, since the Tableau users are saying that it's creative.

It's quick.

It's easy.

Informative and fun.

The words that were mentioned four times -- and I couldn't have defined the Tableau better

than that -- is that Tableau is powerful, professional, and visual.

And the last one that was mentioned the most, is saying that Tableau is interactive.

So I work at Enterprise Business Intelligence Team,

and we are working on a decision support initiative project.

And as a part of this project, we have built a few dashboards.

And you will see -- in the next few minutes, you will see some of these dashboards that are --

that have interactive and drill-down capabilities.

And visually appealing, and informative, and powerful and flexible and so on.

So let's get started.

Here is the dashboard that has interactive and drill-down capabilities.

Again, it is the snapshot of the Tableau dashboard and the slide, so it may not be as interactive.

But it's called Major versus Non-Major.

So what you're seeing here on the dashboard is the analysis of credit hours by the students

in your school, versus the students that are coming from other schools.

So on the top circle you can see, there's a filter.

And you can select a school or some other information.

And when you select those filters, in the first graph you see information that's broken

down for major and non-majors.

So for example, in 2015 there were 42% of students who were majors

in the selected school, and the 57% non-majors.

And so when you click on any of these academic ears, the below graph will filter the data.

And this -- so in this particular dashboard, you're just see a very high level information

of students that are majors versus non-majors.

And if they're non-majors, which schools they're coming from.

So, you know, and you may want to get a little bit more curious,

and may want to know some more information.

So where are these non-major students -- what is attracting them to come to your school?

Or what are the classes that they're taking, so you can decide for the next term if you want to,

you know, expand your class or so on.

So you can select any of these schools.

And as soon as you select one school, it opens a few more dashboards -- a few more visualizations.

And so what you're seeing here is 2015, the students, what classes they have taken.

And then you could even go further, because right now you're seeing the number

of classes that students have taken in 2015.

And when you click on any of these classes,

you can drill down further to see a train of these classes.

So how the data is, you know, how the enrollment is looking like for the last few academic years.

So again, so this was a drill-down capabilities.

They kind of provide the users with very high-level information.

And when they get curious, or they want to --

they ask questions, they can drill down to the information that they are looking for.

So now we are going to move to visually appealing and informative dashboards.

So the dashboard that you're seeing here has a bump chart.

And the bump charts are useful for exploring the changes in rank of values over a time dimension.

So in this particular dashboard, a school is selected,

and what you're seeing here are the departments within that school.

And how the enrollment looks like for these schools --

or for these departments in the last few years.

This dashboard is visually appealing, and it's very informative,

because just in a few seconds you can tell exactly how -- you know, how this -- how the department --

how the school looks like in terms of the departments.

The departments that are pretty steady are the departments

that are fluctuating over the period of time.

So this was a bump chart.

Now, we're going to move to a more colorful, powerful, and creative dashboard.

And it's called a thank you chart here.

So in this particular dashboard, again, it's similar to the first one, major versus non-majors.

You can kind of see that -- so the information that you're looking here, in my --

my students, where they're taking their classes are the --

in my school, you know, where the students are coming from.

And all of the schools are here together.

Right? You cannot probably see the information very clearly at this point, however,

you can select one of the schools, which is one of the colors,

and you can very clearly see the visualization

of how this data is flowing into a different direction.

And similarly, you can also select the other side, and you can see the information from here.

So again, this is, you know, it gives you a lot of information just in one visualization.

And as you hover over your mouse, you can get even more information to just see, you know,

the credit hours, the percentage, and so on.

So all of you see now, that how Tableau is powerful, flexible, colorful, interactive

and has the drill-down capabilities.

So are there any challenges that we're facing?

Yes. There are -- we did come across quite a few challenges.

And one of the biggest challenges is the Tableau's biggest strength, which is powerful and flexible.

Many of you may have heard, with great power comes a great responsibility.

So the Tableau allows you to do a lot of customizations.

You can pretty much do anything to make the picture look the way you want it.

Make it -- you want to look at it.

And to do that, you might have to create a lot of filters.

You might have to create a lot of factions.

You might have to create a lot of calculated fields.

You might have to change the size.

You might have to change the colors, text, and so on.

And, you know, the more you start building,

the harder it gets to maintain those reports in the dashboard.

So the key is to keep things simple and informative as possible.

The second one is balancing aesthetics with the dashboard's usability and design.

So what I mean by that is, as we started working on a DSI project, initially,

all of us came up with great ideas.

We were very creative.

But all of these dashboards looked quite different from each other.

So we decided we need to have some standards.

And so we decided, okay, we're going to use the same canvas size.

We're going to put the filters on one side.

We're going to use some colors and not others.

But pretty soon we learned that those standards did not apply to the new dashboard standards.

So for example, drill-down.

I needed a space on the right-side, so I couldn't put the folder there.

So finding the right balance between the standards

and the dashboard usability sometimes could be challenging.

Excitement.

Well, if you haven't heard the excitement in my voice yet, it is fun.

And one could easily get carried away.

Data interpretation.

So it is human interpretation of a picture.

So is your visualization interpreted correctly?

And to make sure that it is interpreted correctly, you need to provide

as much information and visualization as possible.

Now in the DSI project, we are using a data cookbook to document every single word,

so that information is there to make the right decision.

Or to interpret the data correctly.

Validation, again, it is a picture and the numbers,

so validation as it is, is a complex process.

But in a DSI project we have come up with this process

where we are validating the data at every stage.

But sometimes it could be challenging.

One of the other challenges that is not listed here is the unclear specifications.

So initially when we build a dashboard we get okay, we need to build --

we need the number of students by this and this dimensions.

So as [inaudible] we're going, we start exploring the data in Tableau and we learn,

oh wow, this information looks great.

We go to the functional owners.

Oh, do you think you will need this information?

And they're like, oh yes, that's great.

But can you please add this and that?

And then at the end there's a change of colors.

But [laughs], you know, that is a part of the process where the data occurs,

and the functional owners, they have to work together to define clear specifications.

So now you have heard all the great things about the Tableau,

and some of the challenges that we face.

So you may be wondering, is the Tableau for your team?

Again, you know, Tableau allows you to see and understand your data.

You can connect to pretty much any databases; Oracle, SQL server.

You can connect to an XL file.

And just with a drag-and-drop can create visualization in [inaudible].

So the server dashboard that I built, the initial dashboard was done in 10 minutes.

That's pretty much how long it took to get that data into visualization.

How do I get started with Tableau?

If you go to kb.iu.edu and search for Tableau,

you will find a few articles that would get you started.

So once you build a Tableau dashboard, how do I share Tableau reports with others at IU?

Well, there are Tableau -- there's a Tableau production and test server sites

that are available on us.iu.eduebitools.

And also published dashboards to dsi.iu.edu, which you will hear a little bit later,

more so that you can -- so others can find that.

Tableau information.

Tableau@iu.

So the current Tableau server version we have is 9.3.0.

Just keep it -- be aware that the Tableau client and the server version,

they need to be on the same level for the clients to publish it on the server.

Tableau server upgrade to 10 is coming soon.

I know some of you who are using Tableau are excited about the new features

and the capabilities that Tableau 10 has to offer.

So stay tuned for Tableau updates by subscribing to listserv dstug-l@list.indiana.edu.

Tableau -- so if you are looking for more help on a Tableau,

there is a community -- Tableau community site.

And also Tableau training sites.

And you can get a lot of information there.

[ Applause ]