Robert L. Martin Creative Consulting, New Media Design, Photography, and Printing

Robert L. Martin

Visualization of Big Data

For the last few years I’ve been working on visualizing large data sets to facilitate interpretation. This takes the form of analyzing and exporting data, conceptualizing the important pieces, and constructing a visual metaphor that assists the viewer with interpretation. The process is similar to traditional graphic design, although I’ve been challenged to learn very highly technical tools like R, to do statistical analysis. Most recently I’ve been working to use motion graphics tools to add moving elements to large data sets. This gives viewers additional cues towards interpretation, with speed, arc, color, and direction adding to their ability to analyze information in a visual format. Be aware, this page takes a bit to load, the second flash animation calculates motion tracks for thousands of points of information.

Students and Parents can visually analyze standardized test growth using this tool which visualizes growth and performance using trajectory. For this use, the input fields have been disabled, please click the “Go” Button to see what the output will look like.
Another visualization tool which is currently being developed shows “clumping” of performance data culled from a database. This tool is used to analyze where gaps in knowledge happen over a long period of time. The current data set is approximately 6500 individuals, with 6 variables each.
Some previous visualizations without using motion were used to show performance and growth data to be analyzed by many groups, with differing goals for analysis. In each case they would compare grade and class data, so the visual needed a methodology to interactively change which data was show.
We started with a database, and exported many files like this:  Performance Data
The statistical analysis lead to an initial, not so pretty individual graph:

Each graph was edited in Illustrator, and then flash to get an interactive experience the data teams could work with:

My future research in this arena includes mapping student performance data in ways that emulate the real world. Orbits that wobble when students attendance is irregular, trajectories, and maybe even trees that start to die when performance isn’t up to par. Giving students the ability to interpret their own performance engages them in the learning process, and begins to give them an almost game like interface with which they can use to interpret their own education.

Get Adobe Flash player

Comments are closed.