louisrosenfeld.com logotype

Home > Bloug Archive

Nov 22, 2004: Short Primer on Facets

The recent issue of KM World ran a nice short primer on faceted classification from Endeca's Steve Papa. And not too heavy on the Endeca sales pitch. I especially appreciate the reminder of Joseph Busch's "golden law of facets": "Four facets of 10 nodes each have the same discriminatory power as one taxonomy of 10,000 nodes."

If you're looking for some good examples of navigation systems based on faceted classification, here are some old standbys:

email this entry

Comment: Donna Maurer (Nov 24, 2004)

There are also loads of examples on the iawiki:


Comment: Jed Wood (Dec 5, 2004)

This requirement (according to the iawiki):
*results are displayed as soon as possible*
makes me think of more "realtime filtering" examples, like this one:

Comment: Jeff Werness (Dec 15, 2004)

I've always been a proponent of faceted functionality to enable filtering based on common attributes. My own experience is with Verity's Parametric Search (applicable to both search and navigation). Not oddly, I think the names of Verity's and Endeca's products are directly related to how well-known they are.

While it's probably one of the most effective features a large site can deploy to ease navigation and search, there are obstacles to be aware of, particularly in larger organizations. What I've encountered attempting to deploy this functionality includes:

1. Consensus in standardization of language, for both nodes and attributes/facets. For a very large body of content this can be significant, and the payoff is seldom recognized by much of management and operations. This also demands that decisions are made as to what constitutes a node and what constitutes a facet; don't think this is not a political issue, because it directly impacts levels of visibility for each content category.

2. Divergent goals. Standardization of content deployment isn't necessarily what the publishing teams seek. Whether or not you have a centralized publishing team, the frequent argument is that "our content is different" regardless of what makes sense from user testing.

3. Development teams frequently approach filtering from a coding perspective and how that affects platform stability. This focus is typically at the expense of the larger content issues, resulting in deployments that are ineffective.

4. GUI design. Faceted search and navigation scenarios are very difficult to create without confusing the user. The user's sense of placement within the given task can be obliterated by revealing too many or too few signposts. Epicurious has always been the standard I've pointed to as a nearly perfect example of how to do it right.

5. Complexity of approach. Most non-design staff involved in site development -- even in 2004 -- approach the development of navigation and search from top-down. This is counterproductive in understanding how bottom-up content management is delivered to a user on the site for efficiency and ease-of-use. Additionally, the former is heavily manual and the latter is primarily automated.

All that said, it can be very difficult to sell the idea to management, operations, marketing, design, and technology without giving them all the equivalent of degrees in information architecture -- but when you can, it pays off significantly for both the site users and the business.

Comment: Lou (Dec 15, 2004)

Jeff's comments are spot on; thanks for sharing them!

Comment: Edward (Jun 24, 2005)

Jed, you mentioned the iokio camerafinder demo. They've just launched a destop app for making these things (realtime filtering). www.iokio.com

Add a Comment:



URL (optional, but must include http://)

Required: Name, email, and comment.
Want to mention a linked URL? Include http:// before the address.
Want to include bold or italics? Sorry; just use *asterisks* instead.

DAYENU ); } else { // so comments are closed on this entry... print(<<< I_SAID_DAYENU
Comments are now closed for this entry.

Comment spam has forced me to close comment functionality for older entries. However, if you have something vital to add concerning this entry (or its associated comments), please email your sage insights to me (lou [at] louisrosenfeld dot com). I'll make sure your comments are added to the conversation. Sorry for the inconvenience.