The fundamentals of data visualization: The redux


It's always important to remember your fundamentals. Sort of like a basketball player who practices lay-ups and free throws for hours on end, you need the fundamentals in the midst of the game. Having the skills or knowledge in the heat of the moment - when it counts - is essential.

Visualization of data – whether for discovery or for sharing insights – has its own set of fundamentals. What was interesting to me at TDWI San Diego this week was the introduction to my thought process that we are in the midst of discovering the NEW fundamentals for mobile and big data visualization. We are on the precipice of new worlds, new ways to think about visualization. 

As part of the Business Intelligence R&D team here at SAS, my time the past few months has been spent talking with customers about their mobile and big data needs. How to visualize gargantuan amounts of data? How do you explore it? How do you quickly visualize insights?  On the other hand there are also the questions about how to most effectively use mobile technology. Questions like, how to empower action via mobile devices?  How do we produce optimal role-driven visualization and interactions on the device to speed up the time to intelligence?  

In general, there are established best practices for visualizations – my class at TDWI reminded me of these fundamentals from Edward Tufte. 

Show the data. Induce thinking. Avoid distorting the data. Encourage the eye to see comparisons. Reveal data at different levels of detail. Demonstrate the purpose or objective in the visualization.  And the list goes on. A quote from Rutkowski in the course material summed it up beautifully, that transparency of insight is achieved: "When the user is able to apply intellect directly to the task, the tool itself seems to disappear."

As I sat in my class being taught by Teradata CTO Stephen Brobst and Andrew Cardno and re-familiarized myself with these concepts, I began to think about the ways to apply fundamental best practices as well as types of visualizations to mobile and big data when I return back to SAS on Friday.

I have been doing a lot of research on visualization of big data. There are a lot of great papers and academic research out there -  but I already know that the application of spatial, time series and movement or animation can make amazing things jump out of vast amounts of data.  

The combination of methods is the key. Cardno shared examples of some great projects he has developed in the retail space and  his work prompted a sanity check.  Sometimes I need to make operational decisions, sometimes I am doing rich, green-field discovery, and other times I am sharing insights for another to comprehend and make a decision. Each situation requires different approaches, can be based on 'big data' or not and can be deployed innumerable ways (mobile, dashboard, report, presentation, instant messenger, Outlook, etc.)  Many times the best approach is a blend of several fundamentals.

It's all about context at the end of the day and in the game, you want to pull out your best skills.  If you get a chance to take one of the introduction to business visualization courses at a TDWI World Conference, I highly recommend it. Also if you are interested in visualization or just want to brush up on the basics, check out Stephen Few and Edward Tufte's work as well. Visualization's impact on understanding data is a facet of my job that never fails to be interesting.


About Author

Tammi Kay George

Comments are closed.

Back to Top