I’ve attended many conferences and sessions about knowledge management and SharePoint over the years, and have come to realize that there is a lot you can learn from the knowledge management community as a whole. While broader in scope and essentially independent of SharePoint, knowledge management presents many great concepts that are highly relevant and applicable to SharePoint design and governance. In fact, knowledge management should be a high-level design goal for any information system architecture.
Information warehousing is crucial for recordkeeping, but turning a trove of facts and details into meaningful information is the end result of proper information architecture design. “Knowledge” itself provides meaning and context. For example:
- Information – “I gave my daughter a calendar for her birthday this year. It was purchased at 50 Main Street for $12.95.”
- Knowledge – “I’ve given her a calendar for her birthday for three years; she hasn’t used the last one since she keeps her schedule on her phone.”
The knowledge about the calendar allows me to make decisions and take actions – like asking my daughter if she’d like something else or picking a different gift, for example.
Decision-making is the probably the most important reason to develop knowledge management systems. Few organizations have disciplined processes in place for decision-making or capturing the reasons behind a decision as part of a self-learning process. Getting to knowledge is hard, but it’s not impossible. Insurance underwriting, package routing, and medical scheduling are all good examples of properly designed decision systems. Information architects (and business intelligence designers) should ask themselves if they understand the kinds of decisions – or decision processes – that will use their designs.
With that in mind, the following are six key attributes of an effective knowledge system:
Purposeful and relevantRecords management is sometimes about preserving or "freezing" data and information until it can be applied in the future to create new knowledge. But recordkeeping for recordkeeping's sake is not knowledge management. I could record the weight of each calendar gift I’ve given but, since I’m not shipping it, it’s highly unlikely that anyone will ever need that information again. It doesn’t add to our knowledge. Relevant means more than someone caring about the data at creation. That information should have a clear owner throughout its life, and the lifecycle should reflect its purpose. Separating the “wheat” of knowledge from the “chaff” of data isn’t simple – but that information lifecycle should allow you to permanently exclude information that will never be needed again.
RecoverableWe need to be able to get knowledge back upon demand. In part, we need to recreate the original context of the information. For example, how was it established that my daughter doesn’t use her calendar – observation, hearsay, or her direct statement? Second, as information is analyzed and compared with other information, we need to be able to recreate the context for re-analysis. That means understanding why I was looking at gift purchases. Was it for family budgeting, or to give her godmother a new gift suggestion?
Transferable and impersonalI’m not suggesting that information need not have security. However, knowledge can and should be able to move among multiple appropriate stakeholders, originators and consumers without obstacle. It needs to be persistent without requiring the originator’s personal involvement. If the information as recorded says “Cal Dev BDay Wal 11/1,” it may be obvious to me that I bought a calendar for Devin’s birthday at Walgreen’s on November 1. But anyone else would probably need to ask me what I meant. That’s not usable knowledge. Someone needs to be able to work with my information without my involvement. For effective knowledge management, information needs to be able to be moved from originators or holders without significant changes that require personal interpretation.
RecontextableInformation shouldn't only make sense in its original context. If you can take granular information and rearrange it for new meanings, you can create new knowledge. For example, "95 degrees" requires additional information (it was at midnight on December 25 in Boston, MA) to make it equally relevant in a list of Christmas weather anomalies and all-time daily high temperatures.
RecreateableNot all information is created equal. Some information is more accurate than others. Also, the perceived usefulness of information may change over time. That’s why it’s important to know the source of any single piece of information.
TranslatableIn case I haven’t been clear, knowledge is a higher abstraction than a series of recorded facts. To sustain that high level, the underlying information needs to be able to change storage, formats, and frameworks without necessarily changing the underlying knowledge. The knowledge that John Adams had a more contentious U.S. Presidency (1797-1801) than his predecessor, George Washington, persists whether or not the underlying sources include print encyclopedias, Wikipedia, or a series of Twitter posts.
SharePoint can be a smooth way to multiply and accelerate shared enterprise knowledge. It can also become an unmanaged dustbin of ungoverned content and dark-matter information clutter. Applying some of these concepts to your SharePoint design will help your users share knowledge instead of just data.