The scalability of findability - can IA move beyond web sites?

In the IA community we're fond of findability. It's a simple conceptual hook that lets business grok a key aspect of our work. However, findability also sets some arbitrary boundaries for the practice, and runs into challenges once we move beyond single web sites. Taxonomies and facets just don't scale across the web as a whole, and struggle to be globally relevant in cross-disciplinary enterprises like General Electric.

It makes most IAs cringe to think about automatic categorization tools. However, it's also the inevitable future of large scale findability efforts - no IA superhero can manage billion-document findability from traditional top-down or bottom-up approaches evolved to address site level issues.

Automated classification and semantic analysis is important for people who plan IA careers lasting into the next decade. We don't all need to become Autonomy drones, but it's worth keeping a finger on the pulse. One interesting project that is going to go commercial with Factiva is IBM's WebFountain. WebFountain is also different than many alternatives because the idea is to build a platform for findability - letting other people build modules that tailor the WebFountain base for particular uses. While most of us don't have to deal with enterprise architecture and beyond today, staying relevant in the future will require us to understand the issues at play. Rather than dismissing the advance of machine categorization and semantic analysis, we should be prepared to take advantage of new tools that further findability and ultimately the user experience.