Understanding Managed Metadata Services in SharePoint 2010

One of the most compelling features in SharePoint 2010 is Managed Metadata Service, or MMS for short. Simply put, metadata is data that describes other data. A good way to conceptualize metadata might be to think of a photograph that you’ve taken in the past. The photo is the actual data, and metadata that describes it might include the size of the file, where the photograph was taken and who is in it. In this post, I’m going to focus on how MMS can help you manage metadata, and how, if leveraged properly, it can significantly improve the quality of Enterprise Search in your organization.

Why do I need metadata? Why can’t I just search like I do on Google?

This is the question that comes up most often when I’m at trade shows, conferences, what have you. From a user’s perspective, all they do on Google is enter two or three keywords, and voila – relevant results! What they are not seeing happens behind the scenes, before their query is ever run. The two or three word query is boosted by metadata.

As explained below, you can deliver Google-like search to your users, provided you have the same advantages that Google has:

1. Unfettered access to data repositories
2. Content producers painstakingly tagging content so that it will rank high on Google, Bing, etc.
3. Content consumers unknowingly identifying and tagging the most useful content.
4. Google doing some pretty cool stuff on the backend (Top secret, but much of it is pretty obvious).

So what does this have to do with metadata? Numbers 2 and 3 above are all about metadata. In many ways it is the highest quality, in fact. Starting with 2, human beings are much better at classifying data at this point in time than machines are. On the web, administrators go to great lengths to make sure their content can be indexed and ranked favorably. Google provides a collection of tools to assist admins in complying with their standards.

The point here is that in the Enterprise, most users would never put forth this kind of effort without seeing an immediate benefit. People are too busy. So when you enter that two or three word query, it’s exactly that. The search engine simply does a brute force match of keywords and you end up trolling through hundreds of matching documents; hardly the Google experience.

The 3rd point above is what made Google famous. While in college, Google’s founders recognized that researchers would often reference the work of others if they valued it. Quality research was referenced more often. The “light bulb” moment came when the founders recognized people on the Internet do this as well. They realized that when quality content is published on the Internet, people link to it. This became the foundation of their now famous PageRank algorithm.

Hopefully now that you have a sense of how important metadata is to Google’s success, you’ll take a bit more time to read how MMS can help dramatically improve search for your users.

What is MMS?

Managed Metadata Services is a service provided by SharePoint 2010 that does three things:

1. Provides a central repository for managing an organizations Taxonomy(ies).
2. Provide a central repository for managing an organizations Folksonomy(ies).
3. Provides the ability to syndicate Content Types

Before explaining all of this, let’s get some terminology out of the way:

Taxonomy – A Taxonomy is a collection of related words/phrases that are typically arranged in a hierarchy. The picture below will probably give you a better sense of a taxonomy than me trying to describe it.

Microsoft uses the term Term Set in place of Taxonomy. They use the word Term to describe the words/phrase that constitutes a taxonomy, and they use the phrase Term Store to describe where Terms Sets are maintained and managed. Term Store represents an occurrence of MMS. So think of MMS this way:

TERM STORE (MMS)
     TERM GROUPS
          TERM SETS
               TERM 1
                    TERM 1a.
                    Synonym 1a.

The General Business Taxonomy depicted in the image above represents a Term Set.

Additionally, please note in the hierarchy that I depicted you can associate synonyms with terms which is essential for a good search experience.

Important Point! - Creating/maintaining a taxonomy is not a trivial undertaking. Many people do this for a living. As such, a taxonomy is typically centrally controlled by a Taxonomist, Librarian, or other professionals with deep knowledge in an area of their expertise.

The Internet has shown that social networking is a powerful means of capturing and sharing knowledge. One aspect of social networking is user tagging. Users apply tags to content that they like (or sometimes don’t!). Such tagging is often called a Folksonomy. Microsoft decided to create a new term here as well. In MMS, terms that constitute a Folksonomies are called Enterprise Keywords. Like Taxonomies, these are centrally stored in the Terms Store as users apply them. Unlike a Taxonomy, users can use any tags and are not constrained by the Terms in a Term Set (unless this is desired, but that defeats the whole purpose, right?).

The final capability (labeled 3 above) that MMS brings to the table is around Content Types. Think of a Content Type as a document or list item that has known characteristics (metadata). Let’s say I was a stamp collector and wanted to track my collection in a SharePoint List. I could create a Content Type called ‘Stamp’ with the following required properties (Age, Country of Origin, Condition). When Admins create document libraries, they will often specify a column as being of a particular Content Type, knowing in advance the type of documents that will be stored therein. This enforces consistent property values across the document library.

You might ask what this has to do with MMS? MMS enables the creation of a Content Type Hub, which centralizes the management of Content Types (CT). Site Collections, even other SharePoint Farms, can subscribe to these CT’s enabling consistent metadata across an organization.

Above I described the building blocks of MMS. Now, what can you do with it?

Term Set (Taxonomy) and Term Management

The Term Store enables management of Taxonomy and Terms. Admins can create a taxonomy from scratch, for example. The Term Store gives you the ability to import Taxonomies in a CSV format so you can easily create one in Excel.

An alternative to importing an existing taxonomy is to purchase one. BA Insight has recently partnered with a Taxonomy specialist called WAND, that offers taxonomies across many different industries and domains. As part of this partnership we are offering a FREE Wand taxonomy for General Business needs. Please download it here.

Regardless of what you end up doing, the Term Store enables term management.
The image above depicts some of the operations that can be carried out on terms. Some of the more interesting ones include:

• Merge Terms – Terms that have the same meaning can be merged. One or more of the terms simply becomes synonyms.
• Deprecate Terms – Any term which should no longer be used should be deprecated. This won’t delete the term from documents that have been tagged as such, but it can’t be used going forward.
• Move Terms – Terms can be moved between nodes, or Term Sets.

MMS and Information Architecture

One of the stand-out features of MMS is the ability to distribute the Administration of Term Set globally and locally. Administrating organization’s Taxonomy / Taxonomies is time consuming. They are constantly evolving and require continuous management. To Microsoft’s credit, they recognized that a completely centralized taxonomy is not optimal.

Enter Global and Local Term Stores

As discusses, the Term Store is a central repository where taxonomies, terms, and Enterprise Keywords are managed. This information can then be shared across the organization. Left alone, this would be problematic from an administration perspective. For one group, or coordinating multiple groups to manage a Term Store, would be a challenge.

To lessen the administrative burden on Term Store management, Microsoft introduced the construct of Local Term Stores. A Local Terms Store is a Term Store managed by Site Collection Administrators. What’s fantastic about this is that you can push the administration of taxonomies to the people that have that particular domain expertise.

In the image below, we have the Sales, Marketing, Product, and Legal teams, all subscribing to the Global Term Store (portrayed by the black arrow). Meanwhile, the Product Team (which develops Top Secret products) has their own taxonomy, which is strictly confidential. Both, the Product Team and Legal Team, subscribe to that. The Legal Team also works on confidential M&A activity, therefore must maintain their own Taxonomy as well.

Leveraging Metadata with Search

With MMS in place, and all of this wonderful metadata available, your organization can enjoy many benefits – records retention, metadata driven workflows, etc., but this post is focused specifically on search and findability. Users can expect a dramatic improvement in search by leveraging several features provided out-of-the-box with SharePoint (mentioned below) and optionally, products by third party vendors such as BA Insight’s AutoClassifier.

SharePoint’s Metadata Navigation for Lists and Document Libraries

One of the out-of-the-box features that SharePoint provides is the ability to filter Libraries and Lists. In the image below, the user has the ability to filter on any part of the Term Set. This gives the user the ability to start at the top of a hierarchy and with each selection, see the child relationships of the parent node that was selected. Alternatively, the user can go straight to the child node and filter directly. In either case, a lengthy list can be filtered with a few well pointed clicks.

Another capability that SharePoint provides out of the box is the SharePoint Search refinement capability, depicted below. Metadata is presented on the left of the search result. Users can refine results, what is typically hundreds or thousands of documents, down to a handful by selecting one or more of appropriate refiners.

To Conclude on MMS

Managing metadata is work. Often times, there is a lot of pushback from users because they don’t see an immediate benefit to doing so. But make no mistake though, the organizations that do make this effort will be more competitive as a result. MMS will make it a lot easier for employees to find content and substantially improve search within the enterprise.