Enterprise Architecture, Knowledge Management – Knowledge Representation

I was pleasantly surprised when I came across this definition of Knowledge Management by Dave Snowden.

I like this definition of KM but was struck by how it could equally apply to Enterprise Architecture, in particular when Dave mentions processes and technology, and especially when thinking about the contents of our Enterprise Architecture models.

Back in the 1990’s ‘boom’ days of Knowledge Management, there were a number of definitions of KM – many of which seemed to be tied to tools (hmmm, sounds familiar…) – but the thing that stuck in my mind was that knowledge has to be transformed from the ‘raw’ information about the organisation. This transformation is what makes the knowledge applicable to any relevant scenario, having captured information about specific instances of it. And this transformation of information into knowledge – to some extent analogous to the transformation of data into information – is what never seemed to happen in any of the early Knowledge Management solutions. Apparently, all that was required was an Intranet where anyone in the organisation could contribute their knowledge and then anyone who needs to apply that knowledge just searches the Intranet to find it. More latterly, Web 2.0 was going to solve those problems all over again using wikis and blogs and all the social tools, where people could contribute their knowledge so that others can find and apply it. For some areas, wikis can be very useful if notoriously hit-and-miss when searching.

And that’s the real trick that none of these solutions solve; how do you find the knowledge? How can a search engine make the link between a blog posting about a particular topic to a request for best practice on a process? Without the capability to do this, I don’t think any of these solutions are what I would call Knowledge Management. Without the ability to abstract the contextual, subjective information to build generalised knowledge we can’t apply these valuable contributions to other similar scenarios.

How do we transform the information into knowledge? Well, at its simplest, we need some form of ‘framework’ on which to hang the information that we have to hand, so that we can apply objective definitions to what the information is about and relate it to other information with strong semantics. We can then use this knowledge framework to query – not just search! – the information that has been captured from many different sources and in a meaningful way.

Wait a moment, that sounds just like an EA meta model. It certainly does and that’s when I start to wonder, is it that actually Enterprise Architecture and Knowledge Management should be effectively the same thing – it is just that their current guises are particular perspectives on the knowledge base? For me, Knowledge Management has to be more than a bunch of Intranet sites and wikis and similarly, Enterprise Architecture models have to be more than a bunch of diagrams and pictures if we are to truly apply the knowledge that people are trying to capture.

The meta model gives us a framework for representing knowledge about the enterprise that defines the types of information that we need to capture and how this information relates. The resulting model of the organisation enables us to apply the knowledge to the particular problem that we are dealing with right now, whether it is a large scale, strategic issue or an fine-grained query from the operational or tactical agenda.

And this is what Enterprise Architecture is really about. EA is not the modelling. Rather, EA is about the application of knowledge about the enterprise (ideally, represented in a model, because it rapidly gets very complex) to the problems at hand.

In taking a knowledge representation approach to EA and KM (and I believe we should), is Enterprise Architecture modelling the same as Knowledge Management?

Contact Us