. A brief synopsis of the report follows.Synopsis: The
Prototypical Knowledge Management Organization. Dr. R. Kaplan. ACCSYS Corporation. May 2000.
This report focuses on the "design" of a prototypical knowledge management organization. The premise of this report is to design an
organization and describe the positions in it that would support a complete knowledge management effort. The report presents the organizational design, describes 11 positions, and provides detailed job descriptions for each of
these positions. The content of the report could be used as the basis for a company's own knowledge management effort, as a reference for the necessary positions and the specific requirements for these positions. In addition,
the report also describes a "starter" or "seed" organization that would serve as the beginnings of a knowledge management group within a larger organization.
+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
Who Needs Knowledge Management?
July 2, 2000
More often than not these days I wonder why I am doing
this? Why am I spending time to come up with ideas that can be turned into these monthly columns? I especially wonder when I receive email messages saying, "PLEASE UNSUSCRIBE ME." When each and every newsletter has instructions for
how this is done. It scares me to think that so much information is ignored.
But I am reminded of how important knowledge management is when I see examples of "non"- knowledge management and wonder about them. These
examples fuel my passion for knowledge management. I think we should teach it to kids in elementary school.
Case in point. Just the other day I read an article about a technical area of computer science called algorithms.
Algorithms are pretty important in computer science, as it is from algorithms that programs are created. The article was of interest to me because the author proposed a new way to classify algorithms that he claimed to be "better"
than the present means of classifying them.
Momentarily digressing let me tell you why this subjct is near and dear to my heart. For many years I taught a course in algorithm design at a local university, so when I see an
article like this my interest is piqued. Alas, the article did not live up to my expectations.
When I ask the question, "Who needs knowledge management," I answer, "The author of this article."
Why? Because had he used some of the principles, he would have realized many of the things he overlooked while writing this article. Some would call omitting these things bad scholarship, and perhaps good scholarship is part
of knowledge management or vice versa.
Knowledge management creates a fabric of related knowledge including principles, facts, information, rules of thumb, etc. The fabric becomes a very powerful thing when it enables
people to understand a particular domain as a continuum. So although algorithms and the classification scheme that has been used and the one proposed by the author may represent Algorithmic science (assuming there is such a thing),
they do not take into account other, relevant developments. Let's consider one additional aspect that might enrich the classification scheme and make it more relevant to present day activities in computer science.
There has
been extensive development in the area of object-oriented paradigms. Object-oriented paradigms are related to algorithmic science by virtue of the patterns represented by their object classes. Patterns are another kind of
algorithmic description. And patterns represent abstractions of the processes that will be defined in the object-oriented paradigm. Likewise, when we describe algorithmic types like divide-and-conquer and Greedy-methods, we are
defining patterns. To consider any classification scheme for algorithms without considering object-oriented patterns leaves out an extremely important consideration when proposing any new algorithmic classification scheme. The
author of the paper made no mention of object-oriented paradigms. But what does this have to do with knowledge management?
Let's suppose we could map the potential domain of algorithms. What would this include? It would
include many different type of algorithms, the current types of classifications of algorithms, and algorithm construction methods among other things. As we began to map the domain of potential algorithmic knowledge we would get a
complete sense of the different possible classes of algorithms.
Although this example only represents a small portion of the algorithmic domain it was clear to me that the author of the article did not do this. He took a
limited view, identifying some of the more useful algorithmic types with very little evidence of their relevance to the new classifications. I contend that if he used a more knowledge-based approach the ultimate result would have
been more relevant and more useful to present day algorithmic science.
Classification is a knowledge management task. A classification scheme not only requires specification but also testing. How do we know if a
classification scheme can be used to describe what the classification scheme was intended for? We test. Testing verifies the efficacy of any classification scheme. If a process does not test well, then the process is probably
requires adjustment at best and replacement at worst. Writing a classification scheme is not just about defining classes. It is about creating valuable knowledge and managing that knowledge. The conclusion of this observation is
that perhaps that which we call knowledge management is something we all need (or should all do) on a regular basis to take advantage of the rich fabric of knowledge.
+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
The knowldgWORKS News is written in its entirety by Randy Kaplan and edited by Harriet Trenholm. Suggestions for the newsletter should be sent to
