The “four layer” model applied to unstructured content

In my book, Information-Driven Business, I introduce a four layer model for information.  You can also read more about this model in the MIKE2.0 article: Four Layers of Information.

As a reminder, the four layer model provides a way of describing information in every organisation.  The model explains how information is consumed (layer 1: metrics), navigated (layer 2: dimensional), held (layer 3: atomic) and created (layer 4: operational).  Using this model helps to organisation to understand where it is overly dependent on staff or customer knowledge to manage information at any of these layers (such as summarising to report, or slicing-up in spreadsheets to answer questions).

Some people have commented that the descriptions I use in the book, and are used in the MIKE2.0 article, are geared towards structured data.  To help readers understand how the model equally applies to both structured and unstructured data, the following definitions of each layer may help

Layer 1: Metrics
For information to be used for management decision making, it ultimately needs to be summarised into a score or metrics against which “good” or “bad” can be defined.  This is the same regardless of whether we are talking about structured data or summarising a collection of unstructured content.  The metric for documents could be as simple as a count (for example, the number of policies) or a combination of factors such as the number of processes covered by a particular type of policy.

Layer 2: Dimensional or Navigational
While formally described as the dimensional layer, it is perhaps better described as the way that the organisation can be navigated.  At this layer we are talking about structuring the content in way that we can find in a systematic way (via a taxonomy).  It is from here that metrics, such as a count of policies, can be derived.  It is also from here that we go to find content in its general form (“get me all procedures associated with disaster recovery”).  For instance, in this layer policies can be cross referenced against each other.

Layer 3: Normalised or Atomic
In the unstructured sense it is better to use the term “atomic” for this layer which contains the content in its original form reference by the event that created it rather than a business taxonomy.  This layer is often handled badly in organisations but can be as simple as recording the time, author and organisational hierarchy.  It can also be aligned to business processes.  For instance, in this layer, policies and procedures should be fully formed but only associated with the scope that they are covering.

Layer 4: Operational
The fourth layer is the front line and refers to the situation and technology context in which the content is created or updated.  Examples include: social media, documents on network drives and email within the inbox of the conversation participants.  For instance, in this layer, policies are created (maybe in many parts) but have no context.

Comments

[Steve]

Thanks for reminding me of this model - it is relevant to some current work I am doing. In my case however, I think that you missed the most actionable part of your MIKE 2.0 article. So I will repeat the section:

"During the first implementations, it is important to remember that the role of the architecture is to record data that is already being created rather than try to solve every content problem at once. ... By simply encoding in databases data that is being created in spreadsheets, the organisation is gaining an enormous amount of protection and opening the door to productivity improvements as teams begin to share complex data management processes."
http://mike2.openmethodology.org/wiki/Four_Layers_Of_Information

That is great advice and I can tell you from direct experience that 'capturing' this data is an excellent way to bootstrap a data quality/metrics/analytic effort at relatively low cost. Especially as it does not necessarily involve the purchase of expensive specialist solutions for doing things like MDS, etc.