In the knowledge management world, a common starting point is discussion of "hierarchy" of data-information-knowledge. And some add "wisdom" to the end of this chain. I've written about it too. But the model breaks down pretty quickly, when one applies it to other work on how we work.
David Snowden has mentioned his distaste for the model, particularly when "wisdom" creeps into the picture. But he's come back from a conference with a slightly different model from John Poindexter that seems to make more sense. The model does a better job of the iterative nature of D-I-K and then adds Options and Action, which starts to sound like Boyd's OODA Loop (observe, orient, decide, act).
Snowden has a copy of Poindexter's slide , which shows a fairly linear model. I agree with the comment that the model needs to look more like a loop. Patrick Lambe comments that even this kind of linear progression is too structured: data, information and knowledge build upon one another in ways that a hierarchical or linear model just doesn't fit. That's one of the problems with the usual D-I-K(-W) model: it's an attempt to show a hierarchy of "meaning" or "importance" full stop. When we try to add actions and loops to the model, it founders.
Here's my rough take on a more circular / iterative version (click for full image). Data, information and knowledge are pieces of a loop that involves analysis, sense-making and path-finding. Once you've done enough of that, one can formulate options, take action and have a look at the results and process some more. (There is trouble with this model too - the boxes should probably be consistent in showing things or activities.)