Mapping the Future
by Noah Robischon
February 4, 1998
Five years ago, Harlan Hugh dropped out of the University of Toronto, an 18-year old drifter. Then, one day, when his brother needled him about his future plans, Hugh shot back that he dreamed of building "a computer that works the way your mind does."
Spurred by his brother's challenge, Hugh set to work assembling The Brain. It takes the directories and subdirectories of your hard drive and transforms them into a graphic model that looks more like the machinery of the mind than the machinery of Microsoft.
What this offers the Net's data gluttons is an welcome antacid for the heartburn that comes from information excess. Who among us hasn't plugged "pussy" into a search engine and been shown a bunch of offensive feline-lover web sites? The solution may lie in so-called content visualization schemes. Not only is it an alternative to a web overflowing with data, but it also -- finally! -- frees the Net from its hierarchical constraints.
Perspecta Inc. is probably the best-known company hawking content mapping for the web. Encyclopedia Brittanica hired the firm to create 3D versions of the encyclopedia and dictionary so that people can swoosh around the words instead of clicking and sorting through them. After calling Perspecta's director of product marketing, David Clarke, I learned that even Pathfinder will soon be implementing "SmartContent" across its entire site (thanks for keeping me in the loop, guys). Old fashioned links will become "Perspectalinks" that, when clicked on, pop up a new window or layer with navigation for related information.
The server-side of the system will also be able to log usage and create navigation based on popular routes, serve up targeted ads or point people toward related merchandise. "Instead of trying to figure out what a user might have been interested in we capture concepts that users navigate," says Clarke. "We know exactly what a user was thinking."
Sounds ominous, but useful -- maybe even for users. It worked for Mark Tinkler, who became intrigued by visual mapping while studying architecture at Carnegie Mellon University. He rejiggered the architecture department web site so that it tracked usage and created navigation based on the most-travelled pathways. Seven months ago Tinkler left Razorfish, the design company responsible for The Netly News' current look, to start his own company, Plumb Design, and help clients visualize their web sites with ThinkMap. For an example of a "Mapplet," check out the Visual Thesaurus.
Now Tinkler is helping to create an exhibition for the Smithsonian Without Walls called "Revealing Things" that will launch this month. "The collection is about material culture. The idea is to get people to understand the relationship between items," says Tinkler. "Instead of having to decide whether the jeans go in the '50s section or the rock and roll section, ThinkMap allows you to navigate both."
There's plenty of other interesting work taking place with information visualization, InXight and Semio are two of the more notable companies. The field has actually been around for quite some time, and was pioneered mainly by Muriel Cooper and Edward Tufte. Tufte's new media following rivals that of Tony Robbins; Cooper was a founding member of MIT's Media Lab who passed away in 1994, but her influence is clearly evident in the current trend (Perspecta's founders were her students).
But before you join the concept visualization stampede, be warned that it doesn't work for everything. "When there is no real content -- e-mail, Internet junk and so on -- well you can build maps but the results... it's always garbage in garbage out," says Semio co-founder Claude Vogel. His company is moving away from its original plan to be a visual search engine, instead focusing on mapping scientific data for corporate intranets.
Like the fabled "Push" concept, visualization works for some things better than others, but is equally susceptible to over-hyping. "You have to separate out particular visualizations from whether visualization is a good idea," says Judith Donath, a professor at MIT's Media Lab who is famous for creating the first web-based postcard application. "Some of the problems we'll see is that people will implement the technology without understanding whether it's the appropriate one for what they're trying to show."
Donath is studying community visualization -- interactive information spaces that let people connect through them. One of her biggest obstacles is simply the different definitions of community. For some groups the relevant information is who's online and for how long. For others, it's a map of the different discussion groups to which the members subscribe. Still others are interested in visualizing the social patterns within a web site, much the way Tinkler did, in order to make the place seem more inhabited and alive.
The same diffuse definitions plague the word "content." To some sites it's news; to others it's corporate information or scientific data. What's more, content mapping is a big commitment for both the publisher and the user. The dependability of hierarchical data could be replaced with a chaotic information asteroid field that users are forced to fly through. Even The Brain has its shortcomings -- what happens when you forget something?
Of course, Hugh, a self-taught programmer who claims never to have taken a computer class, is thinking ahead of these problems. Although a shrink-wrapped version of The Brain will be available in the next four months, Hugh is already massaging the next version of The Brain to recognize speech and study your habits to rewire its architecture while you sleep. A worthy goal, but we can't help wondering if it's just a pipe dream.