Let Your Eyes Do the Data Analysis
June 15, 2011 2:52 PM PT

By Rob Knies
Microsoft Research

In today’s data-rich world, people whose work revolves around making sense of it need all the help they can get. Providing that assistance is the goal behind another of the demos being shown during D.C. TechFair 2011: Interactive Information Visualization.

The project is an amalgamation of four separate visualization techniques crafted to enable people to interact with large amounts of data in an intuitive, visual manner. All the systems are interactive, enabling users to rearrange the data so they can best visually detect what is occurring.

“The demo,” explains Gonzalo Ramos, a scientist on the Bing Mobile team, “is composed of different experiences, which are all described best by the global term ‘interactive visualizations.’ The gist of the general concept is that we create tools and systems that express data visually and allow your vision to think, your own brain processes and eyes to detect patterns in the data. We assist you, as opposed to impose on you what the data has. We divide the demo into four different examples of an approach to data analysis.”

One is iSketchVis, which applies familiar collaboration techniques that enable users to sketch charts and explore data visually, whether it be on a whiteboard or on a pen-based tablet computer.

“We explore the ability to use interactive sketching,” he says, “as a way to interact with a graphing system so that a computer and a person work hand in hand to sketch out and manipulate data to view patterns and trends.”

Example of an Euler Diagram
An example of an Euler Diagram displaying relationships between individuals and groups.

Then there are a pair of related visualizations, one called Euler Diagrams, the other LineSets. Each provides a way of understanding the intersections of discrete sets of data.

“We have a series of visualizations that deal with displaying sense and the relationship of objects in terms of membership and hierarchy,” Ramos says. “Euler Diagrams are a generalization of Venn diagrams, and LineSets is a more extreme generalization of the concept of sets.”

The fourth component of the TechFair demo is NetCharts, which enables the analysis of complicated data sets via simple charts that display and offer exploration of aggregated data.

“NetCharts allows you to explore data and the connections between elements,” Ramos says, “to go from the very general to the very specific, and pivot the data in many different ways, such as tag clouds, bar charts, and grids. All the interactions can be done with the data in a very direct manner, as opposed to using cumbersome menus or cognitive infrastructures that are arbitrary and unnatural to people.”

Each of the visualization techniques is guided by a simple goal: interacting with the complex in a more natural way.

“What these particular examples are trying to do is to help you manage data and information complexity,” Ramos says. “I’ll give you an example: search on the web. You start with a data set of tens of millions, and you narrow it down to a list of 10 links. You just did a data reduction. You went from very general to very specific, and you can use normal means to interact with the data.

“The spirit here remains somewhat the same. You use these tools, which are all complementary, part of a tool set that helps you see things in a different light or a different perspective, and you get to a point where things are manageable. If I give you a spreadsheet of 10 million items and then give you a spreadsheet with 10 items, which one would you choose?”

If the demo appears to be all about making things easier for those who have to wrestle with the burgeoning amount of data available these days, Ramos prefers another descriptor.

“I would rephrase that in a way that seems constructive: directness,” he says. “It is about reducing the distance between you and what you want to achieve. It’s about going from noise to signal, rapidly, so you can be effective and good at what you do.”

One more thing: Ramos might not be a member of Microsoft Research, but he is quite familiar with it, and he provides an interesting perspective, given that the owners of the visualization components of the demo, Nathalie Riche and Bongshin Lee of the Visualization and Interaction for Business and Entertainment group at Microsoft Research Redmond. Neither was able to be in D.C. today.

“Microsoft Research is a pretty interdisciplinary and collaborative environment,” he says. “Most of the time we work in teams. I had the privilege to collaborate with Nathalie Riche on the LineSets project, and I’m friends with Bongshin Lee. They asked me if I could step in and help with this, and I said, ‘Why not?’ I love the work, and I think that it needed a voice.”