Making the Web More Graphic

When the Web first boomed in 1995, most people were amazed by both the hyperlinks that allowed users in a single click to leap freely from one context (such as the Civil War) to another (like today's weather forecast in Madrid), and the graphics. The links and graphics of the Web transformed the way in which millions of people—and particularly students under 21—view information.

Now some software developers would like to transform the Web—particularly the way in which users navigate it—into something even more visual. Instead of a Web that still relies heavily on text—on the lists of hits we receive from a search tool, for example—these software producers want us to visualize the Web as the branches of an enormous tree, or a building divided into hundreds or thousands of connected rooms.

So far none of these schemes to depict concepts and relationships in a concrete way has caught on, and most of the companies involved, such as Inxight (www.inxight.com) and ThinkMap (www.thinkmap.com/projects), aren't involved with libraries. Antarcti.ca, however, at www.antarcti.ca/ (the dot-ca stands for Canada, the company's home country), wants to move into the library market. It has developed a method of "mapping" library collections with the company's Visual Net software. Tim Bray, Antarcti.ca's CEO and founder, says that the company's mapping system turns a large database, including a typical library catalog, into a visual "map" of a collection. Bray says that a typical catalog searcher will often get back lists of hundreds of titles, and that it isn't easy for many users to see how those titles, and the concepts behind them, relate. Bray wants to bring Visual Net to public and school libraries. Antarcti.ca is talking to several large automation vendors, hoping to incorporate Visual Net into online catalogs within the next year.

The company's Web site features a graphic version of the U.S. National Library of Medicine's PubMed database, developed at the request of the University of Texas Southwestern Medical Center. Traveling into the Visual Net map is fascinating; topic areas of the database become large, irregular shapes that look like part of an alien building plan, but are labeled "Chemicals and Drugs," "Diseases," and other broad topics. Users who enter these topic areas will see they are divided even further, and each individual document citation is marked with rings and symbols that identify, for example, how new an article is, and whether it's in English.

"A large proportion of school and public libraries are cataloged by the Dewey Decimal System, which most users don't understand," says Bray. "We can turn call numbers into descriptions and graphics." More intriguingly, Bray says that he could also create a map of a library's floor plan, directing users to the exact location in the building where a book may be found.

But some librarians are skeptical of graphic catalogs. Jerry Kuntz of the Ramapo Catskill (NY) Library System says, "These visual navigation systems have been popping up and disappearing almost since the Web started, but none has gained popularity. Doesn't that suggest that graphical outlines... of semantic concepts aren't very intuitive?" But others want to see vendors make catalogs easier to use. Joyce Latham, a doctoral student and teacher at the University of Illinois, wants vendors to pay more attention to existing research, to do more research, and to look at experiments that point the way to "good interface design—for any user." Perhaps Antarcti.ca will draw a map that vendors can follow.