I've been hanging out at the FUMSI Forum recently and I noticed an entry: 'Top Shared and Read Articles in Manage'. It seems James Kelway's 'Designing user-centred taxonomies Parts I & II' is still Top of the Pops, over 18 months after it was written. Re-reading it I can see why - it's full of practical tips for making taxonomies user-friendly. But even following James' advice, taxonomies can quickly become bloated and awkward in comparison to the leaner, more agile alternative: facets.
Part 1 of this article - yes, if it's about taxonomies it just seems to need two parts - introduces the concept of facets and the power of faceted navigation and search. A later article will aim to train any of those lumbering Sumo-type taxonomy trees you may have into the fighting-fit warriors of Samurai facets.
Taxonomies
The word 'taxonomy' has now entered into common business parlance and has, as is the way with these things, adopted a more simplified and generic definition. It pretty much describes any kind of controlled vocabulary regardless of hierarchical structure.
In this two-part article I really only take issue with the portly taxonomies that are still very much de rigueur in our organisations: large multi-level taxonomies with numerous branches that, when visualised, often look like our Windows Explorer folder structures or 'file plans'. As a good friend of mine likes to point out: 'Left to their own devices users, faced with extremely large collections and no help from skilled and experienced information professionals, will start making hierarchies of folders, be they made of cardboard, or be they made purely of electrons'.
How did we get here?
It's not our fault (no but really, it's not). We live with the reality every day. The reality of paper: the physical documents on our desks and on shelves, in folders and in filing cabinets. It is a constant reminder that physical things can't be in more than one place at a time (well, unless you like to keep multiples of things). As David Weinberger's great book "Everything is Miscellaneous" reminds us: the digital world isn't like that. 'Instead of having to be the same way for all people, it can instantly rearrange itself for each person and each person's current task'.
It's hereditary too. We have inherited the organisational tools and systems from this 'one document: one space' reality. Hierarchical taxonomies and folder structures are not just used because that's what people know, but because the majority of tools we have to hand support them, and not facets. The file plan, after all, is easy to create in Windows Explorer. Everyone can do it, no extra software is required. And now we have document and web content management systems and intranet directories that make the mistake of thinking the file plan/filing cabinet metaphor is the best way of organising and finding our digital documents.
The seemingly never-ending reign of taxonomies is in fact not because of their supremacy in organising physical documents, but because we've had to submit to their ongoing influence in the tools we use to organise our digital world too. But where tools do support it, faceted classification is better.
Facets 101 - A Reminder
Faceted classification can be tracked back to S.R. Ranganathan who created the unfortunately named "Colon Classification" which identified five key dimensions of any tangible (or indeed intangible) thing. These properties can be remembered with the uneasy acronym of PMEST or with the five more familiar questions:
Who? : Personality (ie. Subject/Topic)
What? : Matter (ie. Form/Physical Properties)
How? : Energy (ie. Process)
Where? : Space (ie. Location)
When? : Time (ie. Date)
(And now perhaps also the sixth dimension of Why?, as discussed in John O'Gorman's Q6 article.)
The idea is to create at least one list of agreed terms (that could be hierarchical or contain other simple term relationships) from each of these dimensions. Terms from these lists are then used to tag/classify to create a rich description of the item and its 'aboutness'.
The power of facets
Epicurious.com (http://www.epicurious.com) has had a faceted browse system for years and is still one of the most popular recipe websites on the Internet.
By tagging all of their recipes with at least one term from each of their 8 lists (see figure 1), users of the system can decide where to start looking for a recipe for tonight's meal. They're at work, they know they can pop to the shops on the way home to pick up what might be missing from their pantry, but what to have? The weather's looking like it might hold so perhaps something barbequed? Others might need wheat-free meal inspiration for their friends that are coming this evening.
Fig. 1
The facets provided by epicurious.com allow users to start their search with whatever makes sense for them at that moment in time and the results are filtered as they berry-pick their way around the site (see figure 2).
Fig. 2
Compare this with a taxonomy tree (see figure 3).
Fig. 3
The starting folders provided - say 'Cuisine' - force people to begin in only one way. But what if I'm looking for all barbequed food? Do I have to look in every Cuisine branch to know I've seen all recipes for barbequed food? Second, we have to shoehorn some categories together that don't reflect reality (so wheat-free dishes are in the Dietary Considerations folder, but won't some of the desserts be wheat-free too?). Finally, if there's a new food trend, like 'Raw', it means we have to change a lower level concept into two distinct categories, and then we need to dive into every branch to make that change.
Benefits
This is not to say that there's no room for big taxonomy trees or file plans in our organisations; indeed until our tools can support multiple dimensions of classification in the Samurai way there's only so much we can do. (Although Part 2 intends to show some of the preparatory Samurai training that may be undertaken.)
The allegorical benefits of having a mobile Samurai warrior force of facets, over employing a couple of bumbling overweight wrestlers are really those listed below.
Speed
Taxonomies are slow and lumbering because they are often big, monolithic and reflect only a single view of the world.
Facets are fast because they allow individuals to navigate using the mental model that works best for them.
Precision
Taxonomies can be very specific, but often at the expense of precision or accuracy, since you have to bend reality to fit the structure.
Facets can be both precise and accurate because the structure is so flexible.
Adaptability
Taxonomies are not easy to adapt because a change in one branch can affect other branches and structures.
Facets are much more adaptable as whole new facets can be added or changed without affecting the other facets.
As with the Samurai however, facets are not faultless, but as David Weinberger states: 'Suppose that now, for the first time in history, we are able to arrange our concepts without the silent limitations of the physical. How might our ideas, organisations, and knowledge itself change?'
Kate Simpson is a freelance information & knowledge architect currently travelling between the UK, Canada & New Zealand (and plans to emigrate to Canada in 2010). She's spent the last 10+ years turning a legal education into a passion for complex information spaces (law firms, publishing, government, etc). Through her company Tangledom, Kate tends toward the obsessive when exploring the different ways we can make information and knowledge easier to find, use, share and manage within our organisations.
The FreePint Family is a family of resources to help information workers be more effective, raise the value of information in their organisations and contribute to success.
'FreePint... provides most of my professional development because it won't come through work and [other resources] just don't cut it.'
FUMSI Forum: Do you have a research question? Post it to the FUMSI Forum, where professionals share Q&A and useful tips on how to Find, Use, Manage and Share Information. It's free.
![[Reuse options]](http://license.icopyright.net/images/icopy-w.png){kind=small}
I've been hanging out at the FUMSI Forum recently and I noticed an entry: 'Top Shared and Read Articles in Manage'. It seems James Kelway's 'Designing user-centred taxonomies Parts I & II' is still Top of the Pops, over 18 months after it was written. Re-reading it I can see why - it's full of practical tips for making taxonomies user-friendly. But even following James' advice, taxonomies can quickly become bloated and awkward in comparison to the leaner, more agile alternative: facets.
FreePint Shop:
VIP LiveWire:
