Yesterday, I addressed a blog post, or repost by Jeff Carr, on the Early Site regarding myths about taxonomies and SharePoint and I shared my thoughts on Myth #1. Today, let’s talk about Myth #2 – Taxonomy is best left to the Projects Technical resources.

Interesting perspective from a couple of directions:  First is leaving IT out of the implementation of the taxonomy – big mistake in my book. Those guys run the backbone of the infrastructure and are keenly interested in making certain the taxonomy works well on all levels. Why fight with the existing in house professionals? Adding yet another system is a problem without the cooperation of the IT team. Better to build bridges than bring in something that no one in IT will support. That said I do think there is a need to insure that the taxonomy supports the business customer side of the organization. The taxonomist is not likely to be part of the IT staff and most of the taxonomists I have met are stretched to understand exactly how the taxonomy they build gets implemented on the search side or even linking the terms to the information objects they have built it to support and find. What they DO understand is the content they are building the taxonomy for. They create a conceptual model of that content and then a knowledge organization system (KOS) to support it. That may be a taxonomy, could be a thesaurus, might be an authority file or two, could even become an ontology when there are systems to really support search with an ontology.  Certainly we are beginning to see a lot of mash ups and other forms of linked data which I believe will be the next big thing.

The second perspective I found here was that taxonomies are for business transaction, and records management. The world of taxonomy implementations is much broader than tagging things on a web site for e-commence. I do find an increasing number of people using a taxonomy in conjunction with other data to automatically suggest retention schedules for large amounts of content on shared drives. This is a great way to pare down the data stored and can be reliably automatically applied. The entire scholarly publishing community is moving quickly to embracing taxonomies on websites as linked data to show journal articles, grants, upcoming meetings, etc. all served up to the user interface at the same time the item the user was searching for appears. Even related articles based on the taxonomy (rather than purchasing history) can be provided in the user interface as recommendations for further reading or purchase.

Museum collections are finding that taxonomies can help in the search of their collections and guide patrons to additional objects of interest. The challenge with the art museum objects is that the little descriptions in the museum (the small blurbs on the wall by the painting, etc.) seldom describe the painting itself. They talk about who else did things like it and the period of art etc but not the time itself. Other than medium, artist and dates, perhaps a title there is little to go on for the initial taxonomy implementation, but we can save that thought for another day.

Marjorie M.K. Hlava
President, Access Innovations

Sponsored by Access Innovations, the world leader in thesaurus, ontology, and taxonomy creation and metadata application.