Faceted Structures for Enterprise Taxonomies

Last week, I had the pleasure of attending the DATAVERSITY Data Governance and Information Quality Conference (DGIQ). I gave a presentation around taxonomy governance and I was excited that there were people with some good questions afterwards. (And to the person who asked what a taxonomy is, I was so glad you asked and I’m sorry I didn’t catch your name. Feel free to contact me if you have more questions!)

Bram Wessel pointed out in his recent post that the fields of Information Architecture and Data Governance have a lot to offer each other. We practitioners in those fields also have a lot of the same pain points. So at the end of my presentation when someone started a question by saying they were dealing with one giant corporate taxonomy which contains every term, I felt that in my gut. It’s a huge mess and really difficult to manage and use. Bram brought up the idea that the enterprise could use multiple taxonomies instead.

As taxonomy practitioners, the idea of using multiple taxonomies for an enterprise might seem old hat, but our friends in related fields may not have the same level of understanding. For that reason, I’m going to go a little bit back to basics here and talk about using multiple smaller taxonomies versus one big one–why you might want to and some of the things you can do with that kind of structure.

First things first: do you need to have multiple taxonomies? It depends on the size of the organization, how the taxonomy is managed, and how the taxonomy is being used. It’s certainly possible that a single taxonomy is meeting your needs just fine. But as the company gets larger and the business has more information needs that can be met by the taxonomy, it becomes much more useful to use multiple taxonomies.

With multiple taxonomies, each taxonomy has a different focus or subject, giving users a number of different views into the way content or data can be classified. This type of taxonomy scheme, which provides a number of different aspects to the content, is known as a faceted scheme. Creating such a scheme involves getting an understanding of those parts of the business which use the taxonomy, and what is important for them to be able to find the content they need.?

Some examples: almost every department in a business probably needs to know what subject or topic a particular piece of content covers. Sales and Marketing teams might also need to know to whom a particular piece of content is targeted. For them an Audience taxonomy is crucial. An asset management team might need a Content Type taxonomy to determine how assets are stored and what business rules apply to it. A Legal or Infosec team might need a sensitivity or access classification to set user permissions.

Faceted schemes can also help make search engines more robust and easier to use, because each small taxonomy can be used as a separate filter. Someone looking for a white paper about a particular product to share with a vendor will find it a lot easier to do that if they are able to filter by content type and product. Some search engines provide filtering on the search results screen and others hide them in their advanced search options, but either way having the ability to narrow down a search helps the user find what they're looking for without having to go hunting through all the results to find it. Having multiple classifications along different aspects of the content also means that in whichever way a user decides to start narrowing down the results, they’re still going to hit upon relevant content.

Some systems can even use these classifications to power automated and/or personalized content delivery, such as intranet portals, product catalog pages, or e-newsletters. Instead of needing time to manually curate all content, it’s possible to pull together relevant content automatically. This makes the publishing process much more efficient without the need to add a lot of additional technical overhead

Of course, getting the full use out of a faceted taxonomy scheme requires the content to be correctly identified and classified. Good governance requires understanding who will be doing the classifications, how the classifications are done, and what kind of quality assurance process needs to be in place. If you're already dealing with a giant list of terms that users have to browse through to find the ones that they need, you might find that creating a governance process for classification ends up saving time and effort that was being taken up by users having to navigate all terms at once.

If you have any other use cases you’ve come across for a faceted taxonomy structure, or stories you have about moving away from one big taxonomy I would love to hear it. I appreciate any comments and feedback you have.

Thanks!

-Erik