Setting the Terms in SharePoint 2010

Community Topic(s):

Keywords: metadata, term sets, MMS, knowledge planner, folksonomy, taxonomy

Current Rating:
(0 ratings)

Tagging is based on two units -- the link, which points to any page within the range of your security settings, and the tag, which names or labels the link. Taxonomies on the other hand answer to absolute values. Tags are self-evident and self-organizing. Taxonomies come with instructions, mainly broader, narrower and related terms.

The big draw for tags is that they’re easy to create and even easier to follow: “As the Web has shown us, you can extract a surprising amount of value from big messy data sets,” observes blogger and classification heretic Clay Shirky.

Formal taxonomies are generally better fits for well-entrenched fields of knowledge where a shared vocabulary denotes a set of precise, static meanings. Science, medicine, and law are three examples of disciplines where common tasks, procedures, and topics have a well-defined boundary of unique values and fixed connotations.

The language of the marketplace is a foreign one – even to many business classification systems. Perception-shapers like ad firms, media titans, and management consults aren’t rewarded for following precedents. They are expected to stretch and bend them. The argument runs, here’s where informal or Folksonomies take over. The knowledge economy runs on the fashioning of ideas – not the production of tangible products. A taxonomy doesn’t handle interpretations or what we do with the things it classifies. Show a taxonomist a verb and you may get a cross-reference – or be referred to a different taxonomy.

Enter SharePoint 2010.

The great thing about Managed Metadata Services ("MMS") is that it forms the speed-of-thought index that gets out in front of a user’s top-of-mind intuitions. Us ECM managers are not handing our colleagues into the awaiting arms of terms bought and paid in the betting parlors of search media keyword programs. We’re inviting our users to try on a label that suits them. They can parade that label in the ad-free confines of an intranet where consumers are actual colleagues and they’re buying rationales and arguments – not goods and services from Ad Word sponsors.

Here’s how that index resonates through term sets:

  1. First we import the query logs to tally the range, frequency and sequential orderings of popular noun phrases (known to our users as keywords).
     
  2. Then we connect the demands placed on content supplies by these search sessions with the terms inventoried by our knowledge producers. Typically that means six variations on the act of managing. What’s being managed, how, and by who is rarely made explicit but we’re most certainly managing something.
     
  3. Finally we cheat and add a smattering of nicknames, entities, definitions, and shorthand for shop lingo – a capsule of time-release acronyms that could choke a horse. Some of this term set cocktail is a homebrew. But few of the metrics, ratios, and models are sculpted from scratch. Many owe their improbable births to distant economic cycles, remote cousins from recent trade conferences, and a witch’s brew of earnest engineering and inexplicable financial instruments.

What we get are vertical vocabularies laced with process-specific operational jargon. However there’s little purity when it comes to packaging a set of capabilities for a different base of customers. Those generalities require more contextual bearings to register a repeatable set of methods for generating established successes in new segments.

For instance it’s not enough to weave every variant of c-o-s-t into the operational structure of a product cycle; especially when that cycle is calibrated to an R&D culture, a supply chain disruption, or carve out opportunity.

That’s why the grab-bag of noun phrases won’t scale to an overlapping set of objectives. That same pharma giant competes in home and personal care. How do we make sure these not-so-subtle distinctions are not lost on the IP we’ve developed in both markets?

I suggest the following dimensions to round out the contextual boundaries of all term set entries:

Term Scope: What is the most common purpose or unifying objective of the IP in question? Is it a sales document or delivery of a contracted service – on the other end of that output: who’s it for in terms of markets or the kind of project we’re documenting?

Term Type: How is the scope being addressed? What’s the wrapping around the delivery – is it a training program, a business process, or something more standalone like a specific framework, patent, or business method? Who’s on the receiving end in terms of the functional units – think horizontals and the job titles of the folks who sign our purchase orders.

Term Class: Classifying the actual term sounds simple – except that it’s rarely done. Is some densely-named government program a key part of a regulatory practice? Better label it as such. Is is a commonly repeated nickname that transcends its more formal legal identity? We may be familiar with the history but those new hires in the Bangalore office may not go back as far. Share that history through term classes.

Report

Rate Post

You need to log in to rate blog posts. Click here to login.

Add a Comment

You need to log in to post messages. Click here to login.

Comments

sunil k sharma

Good for information anaysis

Thanks for this pretty brief Marc. Will appreciate if you can elaborate lil further on Temsets i.e. the tech specs and real time config settings on term scope, type and class wrt sharepoint 2010.
Report
Was this helpful? Yes No
Reply

Marc Solomon

Assuming Nothing

Sunil -- The three classes for defining, contextualing, and reusing term sets are not configurable in SP2010. They're created on the back-end to help community managers like myself do the necessary mediation between folksonomies and taxonomies. As far as I can tell that will remain a one-way discussion for the foreseeable future!
Report
Was this helpful? Yes No
Reply

Chris Riley, ECMp, IOAp

Yes and No

Marc,

This is a good post. But the managed meta-data services is used in several ways, and this methodology is more akin to Folksonomy then it would be for Taxonomy. I would encourage users to sort that out first. When designing a taxonomy it is a significantly different process.
Report
Was this helpful? Yes No
Reply

Marc Solomon

It boils down to findability

Chris -- No doubt about it. Controlled vocabularies are command/control, hierarchical and a whole lot more wound-up than the flat, informal, and dynamic world of folksonomies. The two however are not exclusive. Co-existence works as a root term, user tag, and bumpersticker. The blog post is not about favoring one over the other but tracking the definitions and origins of our MMS indices.
Report
Was this helpful? Yes No
Reply

This post and comment(s) reflect the personal perspectives of community members, and not necessarily those of their employers or of AIIM International