Relationship visualization
A summary argument:
- Continued growth of internet, mobile, and sensor based applications means an explosion of data---a "big data" world.
- Big data requires visualization to provide meaning.
- Data visualization will become an important and large field of design.
- Data visualization will have many sub-disciplines.
- Relationship visualization is a sub-discipline of data visualization.
Relationship visualization provides tools that help people explore relationships between factors in a system. For example, if I want to take out a loan, how much will I pay overall, if the interest rate is x%? What happens if I pay off the loan in y years instead of 30 years?
Until recently, we've had to make do with tables or graphs or work out the calculations by hand. More recently, we've seen special purpose calculators, e.g., mortgage calculators. But it's possible to provide interactive diagrams---relationship visualizations---which support reader interaction with a set of related factors. Move the interest rate handle up or down, see the graph go up or down.
Where things get interesting is when many factors interact and affect each other. For example, total home price, points, percent down, interest rate, length of loan, number of payments per month. All these factors affect the loan principle, size of individual payment, and total paid. The possible combinations are almost infinite. Users may start in different place. Perhaps you know the total you can afford per month and want to find the largest loan you can afford. Or you want to buy a particular house and want to know what's the maximum interest rate you can afford.
Being able to "play" with the variables---twiddle the factor dials---is a way to experiment which is a way to understand or learn. So we might say that a mortgage factor relationship visualizer is a tool for learning about mortgages. And it's more. Learning and designing are closely coupled. Learning involves acting in the environment, experimenting---prototyping mortgages in our example. So we might also say that a mortgage factor relationship visualizer is a tool for designing your own mortgage. And designing relationship visualizers is (in part) about designing tools which enable others to design.
There's also a technical side to the argument. Tools like Flash and Processing have made data visualization more accessible to designers and also offer possibilities for creating relationship visualizations. In theory, one can create infinitely rich environments with those and other tools, but in practice they exist in boxes walled off from the larger environment in which they are presented---that is web pages or e-books (which are also essentially web pages---for the most part).
However, the advent of HTML5, CSS3, and recent changes to the DOM (Document Object Model) create a growing platform on which data visualizations and relationship visualizations can be built and fully integrated into the presentation environment. In this platform users can manipulate objects on screen creating JavaScript events which redraw SVG objects. This combination has the potential to become quite powerful. What's needed now are libraries---building blocks---to make the development process easier.
We've begun to build examples. We invite your comment.