You can’t apply technology and supporting strategies effectively to a resource if that resource does not have a well-understood model. Applying technology to the superficial (or imagined) properties of an un-modeled resource — for example, all the text in the world — gives you a very small, marginal, and unpredictable return on investment. Consider, for example, how we had to adopt standardization of parts and physical metrics themselves before moving to assembly lines.
What are my goals?
- Provide an abstract reference model — ideally in conjunction with co-conspirators — that helps researchers, designers, and software engineers develop new, complementary tools (frameworks, tools, and techniques) for capturing, managing, and sharing practical knowledge. See also, What are the characteristics of a good model for practical knowledge?
- Develop the consistent language we need to talk about these things more meaningfully — starting with names for the stuff we use most frequently: Concepts, Facts, and Insights(and, yes, I am proposing that we use these and other terms in specific ways). See The Insight-Centric Knowledge Model — ICKMOD.
- Identify Insights — and the Facts and Observations that support them — in the Resources to which you have access.
- Make the meaning of Facts and Insights as unambiguous as possible … in human-friendly forms.
- Record those objects and the relationships among them in forms that can be edited.
- Eliminate duplication of Insights (and other nodes and relationships) by making their meaning explicit.
- Provide a consistent model for evaluating the relevance, truth, and importance of those objects — and the relationships among them.
- Support the transition of knowledge capture and transfer from a cost-centered business activity to a value-centered business activity.
- I look at the challenges of capturing, managing, and sharing practical knowledge from the perspective of work and management of that work — not from the perspectives of how we represent human knowledge perfectly and not from the perspective of computer-supported interpretation of natural language (NLP or Natural Language Processing). We can’t represent knowledge perfectly … and we don’t need to. And we need much more than NLP.
- However, we can learn from the thinking done in many different domains and communities, ranging from Knowledge Representation, argumentation, and Discourse Theory to simple computer tools for “concept mapping” and “mind mapping.”
- I approach the problem from the perspective of communication of meaning — a “semantic” perspective, if you will. It’s not about boiling the sea of information in which we find ourselves.
- My short-term aspirations include supporting the transfer of rich semantic information between the many computer tools that are already available to us.
- I believe that clear models for representing practical knowledge make it far easier to connect the tools and practices of knowledge management with the needs of knowledge work.
- Representations of knowledge that cannot easily pass the test of expression in multiple natural languages should be viewed with considerable suspicion.
- I leave it to others to address issues of motivation, cooperation, in “knowledge management” and “knowledge sharing,” etc. Those issues are not irrelevant or unimportant, but they represent a different dimension of the challenge.
© Copyright 2017 Philip C. Murray