Data, Information & Knowledge - The Process of Learning

A key question that frequently arises in Data Architecture is: What is the difference between Data, Information and Knowledge?

It might seem an odd topic but understanding the “Process of Learning” is a key aspect of understanding the separation of these three classifications in order to create a true Knowledgebase rather than a mere Database.

Learning is a process involving a set of discrete states that the Learner (anyone who wants to discover new knowledge) passes through, and a set of activities carried out by the Leaner in order to move from one State Of Learning to another.
All learning starts from a position of ignorance and over time, through a structured process of gathering “data”; understanding it; and testing ideas, aims to gain knowledge about a given subject.

It may seem obvious but without a structured approach the results of the Learning Process are uncertain and, more importantly, critical knowledge required to fully understand the subject area may be missed or omitted entirely.

The Learning States and the activities for transforming from one state to another can be summarised as:



Each State indicates the relative position of a Learner with respect to a particular Learning Domain. The States that may exist are:

• Ignorance is the state of not knowing something about a particular Learning Domain.
  
• Data is basic Learning State where the Learner possesses the raw and unformatted facts and observations that are available or may be collected within the Learning Domain.
  
• Information is the Learning State where the Learner can provide Information about what already exists within the Learning Domain and is able to retrieve that Information in an organised and structured manner. This is the “Body of Truth”
  
• Knowledge is the state that is achieved by having Information that is understood by the Learner. This includes all facts that can be systematically and verifiably derived from the underlying Information that forms the “body of truth” on which a particular degree of confidence can be placed.
  
• Wisdom is achieved when we can deduce or predict the likely answer to a question by recognising it’s similarity to other knowledge we possess without having to re-enter the Learning Domain from the beginning of the Learning Process.

Note: We could also argue that it’s the degree of assertion, ranging from "somwething I've recorded" through “something I believe” to “something I know”, representing the confidence level that the Leaner applies to the Facts that would classify it as Data or Infprmation or Knowledge.

In addition this body of Facts can include “incorrect” facts as well as “correct” facts because the fact that something is definitely not true is also may also be meaningful.

The transformations that take place are:

• “Purpose", given awareness of Ignorance, is the process of entering the Leaning domain. It’s formulating the objectives and goals that are set as the reason for learning. Without Purpose the Process of Learning is unfocussed and directionless. This could be regarded as a pre-condition to learning but, in the case of Human Learning, makes more sense to regard it as a transformation that prepares the Learner for the process of Learning.


• “Classification + Organisation” is the basic process of describing data by classifying the various data-items and organising them into data hierarchies.
  This is the initial task because in order to extract Knowledge the Data must first be placed into a framework that is understandable and well structured.
  
• “Understanding + Interpretation” constitute the analysis methods that are applied to the Information in order to extract Knowledge from the Information that is available.


• “Experience + Application” is the process of “testing” Knowledge by applying it outside of the Learning domain. A possible outcome of this transformation could be a failure of application and the need to re-enter the Learning domain i.e. "Something is wrong and we need to find out what."


• “Hypothesis” is the process of examining Knowledge to formulate questions that require additional Data to be subsumed into the Learning domain.

Of course, the Learning Process potentially forms an infinite loop because there is always at least one more Hypothesis that could be made. The exit condition is a lack of willingness or reason to hypothesize i.e. Purpose has been satisfied and the necessary Knowledge has been acquired.

In addition any individual element within the Learning Domain could be the subject of its own individual Learning Domain and any number of Learning Domains can be combined to form larger Learning Domains.

Of the three transformations within the Learning Domain the “Hypothesis” transformation is a “free thought” process that is very difficult to quantify in procedural terms, i.e. we don’t know enough about free-thought to be able emulate it in an artificial construct so this requires “real people” to formulate the questions that need to be answered.

The other two transformations can however be quantified and, much more importantly, can to a large extent be automated to produce a body of knowledge. That is:

• Data collected from outside the domain can be automatically classified and organised as it is integrated with existing Information already in the domain. Once ot conforms to the pre-defined data quality criteria and can be interrogated in a meaningful way then we have more Information.


• “Understanding + Interpretation” are provide by analytical functions that access the Information and derive Knowledge that is npot inherently apparent from the individual instances of Information.
  
• For example Average Building Energy Rating ratio requires two Facts (the Number of Buildings and Energy Rating for each Building) to calculate it and the result is not directly derivable from either Fact in isolation.

This is “Learning” in the pedagogical sense but it is “Learning” of the Intelligent rather than the Extelligent kind. That is, we may have much of the Knowledge freely available as Facts but in varying States of Learning and, from the perspective of a single person, it exists in many pockets of Knowledge distributed across many individual locations.

From the collective perspective, we want to move towards the Extelligent end of the spectrum with a common and shared understanding of the body of knowledge that transcends the existence of any individual information repository or any one persons knowledge of it.

It is the extelligent gathering together of knowledge into a single location to meet a common purpose that makes a Knowledgebase an important research and decision support tool for understanding one of the most critical and complex issues of our time.