Within the statement “Why is KM so hard” there are two different directions and issues that many organizations face in considering the problem. The first is three reality of What am I actually trying to capture. The second is how do people actually use the information.
Data scientists push data into the traditional data paradigm called DIKM or DIKUM.
· D = DATA without a CONTEXT is meaningless;
· I = Data with a CONTEXT is INFORMATION;
· K = Information in action is KNOWLEDGE;
· U = Knowledge that’s intentionally connected to other Knowledge is UNDERSTANDING;
· W = Knowing what knowledge to put into action, when, where, how and why is WISDOM
This is a very traditional data science view of information. It is missing any number of variances and variables that change the data. We, for example have lot’s of genetic data without context, it isn’t meaningless it just isn’t categorized. I prefer a simple scale for data that is a little easier to evaluate (and one I included with my DLM© system design).
· RD: Raw data, available but may not include schema, context or connection
· PD: processed data, some connections, schema and context
· ID: Indexed data or data with context assigned
· AD: Available data, which is a good system would be the other three combined.
Mine scale is, when considering data to be consumed or provided as part of an Inter-Generational knowledge transfer system is a little cleaner. You have the concept of raw, unprocessed data. This includes all data, all sources. For example there is 110 Zettabytes of data produced by the Internet of Things devices as of 2015. However of that 110 Zettabytes much less is consumed, considered and stored. The rest simply disappears. The next level of data is processed data, that is data we are applying some rules to, but not indexing or moving fully into a search system.
This scale fits cleanly with John Boyd’s OODA loops. Where we are gathering information (observation) we then can apply one of the four filters above to the data we are receiving. Our orientation is controlled more with Raw Data, Indexed data and we can move faster. Raw data can be consumed in a decision making process, in fact Boyd’s OODA Loops were originally designed to improve battlefield decision making. In the area of battle field information you may only get RAW data. Then you have to build a framework around the validity and value of the source for the information. The filters you apply to the information determine the success of the rest of the team, so the right filters are always useful in decision making. The diagram provided is intended to show the relationship between the information and the processing of that information. The value of indexed data is that it is searchable and easily retried. I found this article yesterday, here. Raw data may however come from a source (expert) that is superior to the known good source (indexed) so it is important that the scale not be linear. In the data scientist world they would say an expert is known good or data with context. But there is data we all have that has no context yet we act on it.