Why do we need a new model like ICKMOD?

Why, in general, do we need a model for representation of practical knowledge? Because without a target model for the resources being developed, (1) it’s impossible to apply consistent practices for development and management of knowledge resources and (2) it’s impossible to find, develop, or apply software to support those practices. It doesn’t get much simpler or more compelling than that.

The negative reasons

But here are a few of the other specific negative reasons for developing such a model and for favoring the overall rationale and design of ICKMOD (Insight-Centric Knowledge Modeling) in particular:

  1. In spite of their great contributions to understanding how chains of logical statements about the world can be represented in ways that are processable by computers, Knowledge Representation (KR) experts are not primarily interested in the business/work issue/problem of representing “knowledge” as evident in human communications … and the immediate value of doing so to people. The KR community focuses on construction of programs for unassisted processing of natural language. Many common practices (and supporting technologies) that directly address meaning in real-world situations — including formalized argumentation/discussion and “mind mapping” — are simply excluded from consideration.
  2. The various academic/research communities concerned with formal representation of knowledge don’t even agree on the basics. They are split into competing camps with different assumptions, different goals, and vastly different notations/languages and standards.
  3. Few start with the assumption that meaning precedes language. The demarcation between natural language and representations of meaning is not well defined or maintained (in practice) in KR, in particular. Reverse-engineering principles of representation of practical knowledge from theories of how natural languages “work” introduces complexities and errors that are not only unnecessary but also misleading. Why not start, instead, from very basic principles of how we know and do stuff and how we represent and communicate those processes. Human languages help us understand that, but they are not the most productive starting point.
  4. There is only limited agreement on a shared terminology for representation of objects and relationships that can be used to express meaning. That agreement may be limited to the notion of Concept. And even that agreement is not solid.
  5. Current KR technologists favor rigor. They hold representation of broadly useful, but often fuzzy practical “knowledge” at arm’s length. Unfortunately, most valuable knowledge begins with a rather fuzzy shape — as intuitions, analogies, or informed guesses — not as rigorous formulations.
  6. A corollary to point (5): Except within the practical constraints of application of formal inferencing to human communications, evaluation by people of the truth, value, or relevance of any particular assertion/proposition is not accommodated in any systematic way in the KR community.
  7. The impact of the evolution of new tools for representation (and dynamic visualization and navigation of representations) is vastly under-appreciated. Touch- and gesture-based interfaces are only the beginning.

“You can’t measure what you don’t describe” — the management issue

Aw Kong Koy of Multicentric Technologies — a friend, engineer, and all-round good thinker — appears to have been the first to observe that, “You can’t manage what you don’t describe.” This is a simple but remarkably astute twist on that old saying, “You can’t manage what you can’t measure.”

Both observations are true. Pretending that you’re managing knowledge workers by looking at their outputs (unstructured documents, in many cases) at the end of the month isn’t really management at all. You have to know what they’re doing, how often they’re doing it, the instruments they use, and what their outputs are — at a granular level. And you have to know when those activities change. Describing what people do is a semantic perspective. Attention to semantics makes it possible to describe what people do in a structured way. If you as a manager don’t know what your people are actually doing, then you’re just shouting at the river to try to change its course.

The positive reasons

  1. Quite simply, we need to focus on how we develop, refine, integrate, and apply knowledge … as people. This has not changed, even in the Age of Information and “Big Data.”
  2. All value is created one person at a time. In the case of machines, the designers and builders of the machines are primary creators of value. Secondary value is created by those who control the machine and is owned by the machine’s owner.
  3. It’s right to assume that human knowledge cannot be represented perfectly, but aspects of human knowledge can be represented explicitly with much greater precision and completeness than is possible with natural language … without requiring absolute rigor and/or adherence to the requirement for formal inferencing. Deconstruction of human, work-centered communications into explicit, sharable objects and relationships also makes understanding easier and more complete, provides a method of creating persistent, integrated memory, and makes it possible to point to the meaning of communications and say, “That, specifically, is what I am talking about.” It is sufficient for human business/work purposes that explicit representations of knowledge make it possible to reconstruct true human understanding — as well as to make the judgment that one does not need to understand specific representations — much more easily and more accurately than reading and interpreting unstructured information, which itself is a redundant activity in most cases.
  4. Evaluation (by interlocuters) is an essential, integral part of the development and refinement of human knowledge. This has not changed. At this time, only argumentation/discussion technologies provide some capabilities for this aspect of knowledge representation — and those practices and technologies are typically driven by the goal of resolving specific, “big” problems whose relevance is widely accepted. But “knowledge” plays many other roles, and the value and relevance of any particular piece of knowledge is not known in advance. What’s more, a record of deprecated ideas is essential.
  5. The opportunity to create game-changing applications that address representation and use of knowledge usable by humans is huge. The opportunity to develop both large and small complementary applications is also huge. But it all depends on convergence on a target expressed as a clear, usable model. Neither that model nor knowledgebases created to comply with that model have to be “final.” In fact, the model itself must envision how change may be accommodated. The model does not have to be complete or inclusive or rigidly prescriptive. New stuff will evolve. But you have to start off on the right track.

© Copyright 2017 Philip C. Murray

 

This entry was posted in ICKMOD. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *