Cooperative development of a classification of knowledge management functions

Cooperative development of a classification of knowledge management functions

Philip C. Murray
Editor-in-Chief, Knowledge Praxis

Classifying functions of KM | A starting point? | The vendor opportunity
Functional classification | Why a classification... | How to help us

Sentences, paragraphs, and books never provide sufficient precision to achieve efficient, shared understanding of complex domains of knowledge. Every time you read a treatise about a complex domain, you create a new, slightly different understanding in your own mind. And when two different people read the same treatise, their understandings may be very different indeed. The bigger the domain — and the more information there is available about the domain — the more interpretations will differ, often offsetting the value of added information.

Knowledge management itself suffers from this pandemic problem of our knowledge-based economy. The lack of common terminology and shared understanding of key ideas inhibits a reasoned discourse among practitioners and theorists of knowledge management. It inhibits rapid development of much-needed solutions for the most pressing business problem of our times. It makes business managers reluctant to adopt knowledge management solutions of any kind, no matter how well conceived.

And it recalls the embarrassing paradox of the cobbler’s shoeless children.

The solution: classifying the functions of knowledge management

That paradox can be eliminated only by creating a formal classification of knowledge management business functions — by deconstructing the overall domain of knowledge activities into component concepts and the semantic relationships among those concepts. This effort should be guided by the experience of the knowledge organization community, the classification and library science experts who have proven the value of their work to us all our learning lives.

As information has expanded in other disciplines, a shared classification system has become vital to the continued growth and refinement of knowledge in those disciplines. For example, in medicine the MeSH (Medical Subject Headings) classification system helps researchers, practicing physicians, and medical information specialists organize and find vital medical information. A similar classification scheme has been proposed for music — yet another vast and diverse domain of knowledge.

Although the classification community does not have all the answers, techniques, and tools we need, it is already addressing many of the critical issues of knowledge management, as indicated by the topic of the most recent ISKO (International Society for Knowledge Organization) conference: "Knowledge Organization and Change."

The need in knowledge management goes beyond bibliographic concerns. The benefits in these early years of knowledge management include sense-making, precision of meaning, and more rapid product development and implementation.

So we at Knowledge Praxis are proposing a project to develop such a shared classification system — a road map to the key ideas and the relationships among those ideas in the complex domain of knowledge management — and we invite you to join us, in whatever capacity you feel appropriate.

This project needs a basic model of objects and relationships, a shared methodology, expertise in knowledge organization practices, (with luck) some appropriate tools, and lots of contributors who enjoy sharing their own knowledge in a common effort.

But most of all, this effort needs an advisory board. Good systems of classification arise from review and approval by subject-matter experts. At minimum, we need experts from the realms of cognitive psychology, artificial intelligence and expert systems, organizational management, document management, publishing, and knowledge organization itself.

Name:
Company:
Email Address:
Address:
City:
State:
Zip:
Phone:
Fax:
Your Comments and/or Specific Interests

A starting point?

At Knowledge Praxis we have a model for knowledge resources that transcends specific knowledge-organization tools (semantic networks, thesaurus construction tools) and delivery models. We are willing to share this model and our initial classification scheme — and take our lumps for their inadequacies — as a starting point for developing a mutually acceptable basic model. We are also willing to serve as facilitators of this project and guardians of the resource, using the Media Access Group Web site.

The first step, however, may be to identify work already done in this area. Some of you have already been thinking about this topic — if not as a formal project, then as a real concern that you face when wrestling with the difficult problems of knowledge management. (And if you know of any existing efforts to classify knowledge management functions, please point them out to us … and perhaps save us all much wasted effort.)

Building on his definitions of knowledge (see What is knowledge management?), Fred Nickols recently asserted:

Two factors are thus central to the management of knowledge: the defined body of information itself, and people’s grasp of it. From these two factors, several rather obvious functions of knowledge management quickly emerge:
• acquisition
• organization / specialization
• storage / access / retrieval
• distribution
• conservation
• disposal

Email from Fred Nickols, August 18, 1996.

In a posting to the Knowledge Management Forum, Karl Wiig of the Knowledge Research Institute has loosely categorized some knowledge management applications as:

    1. Knowledge Creation And Sourcing — Build Knowledge through Innovation, Learning, and Importation
    2. Knowledge Compilation And Transformation — Reconstruct, Validate, and Inventory Knowledge
    3. Knowledge Dissemination — Distribute Knowledge to where It Is Needed
    4. Knowledge Application And Value Realization — Use Knowledge to Deliver Products and Services

Nickol’s and Wiig’s categories might be expressed with quite different terms, and the huge pie of knowledge management can be sliced in many ways.

Whatever the motivation or perspective for classifying knowledge management functions, doing so is a highly desirable goal in itself. In fact, describing the component business functions of knowledge management — and their relationships to each other — is the only way we can have some assurance that we are speaking about knowledge management with agreed-upon precision. That precision is sorely needed.

The vendor opportunity

This project also represents a huge opportunity for vendors of knowledge-management products. The future will demand a variety of tools for cooperative, mediated construction of shared knowledge resources, for visualizing the products of that effort locally, and for publishing the results of that effort on the Web and in other formats.

We have examined thesaurus-building tools like Hector Ecchevaria’s MultiTes, semantic nets (including, ATLAS/ti), and KK Aw’s InfoMap knowledge organization tool. We use Corel Corporation’s InfoCentral as a method of data entry and visualization. (Note that demonstration versions of all of these products — including a full working version in the case of InfoCentral — are available at this writing at the respective Web sites.)

Each has its role and its strengths. New tools based on relational database technology could combine many of those features with new functionality and make a more direct connection with knowledge management requirements. There may be many solutions out there that we have simply not uncovered.

In any case, we feel that there are certain minimal requirements for tools that address classification of business knowledge. Those tools should include support for

    • an "unlimited" number of objects.
    • unique identity of objects.
    • definable attributes ("fields") and values for all objects.
    • methods of specifying multiple types of relationships among objects.
    • a "faceted" approach to classification.

We believe that all tools for processing such classification data must subscribe to a common import/export model — most likely a comma-delimited relational database format. The model must not be restricted by a particular technology. It must be able to convey the highest common denominator of information about objects and relationships.

Functional classification as a method of matching problems and technology

Is this the right way to achieve a common understanding of knowledge management functions? There is excellent precedent in the knowledge organization community telling us that functional classification is the way we should approach this sense-making process:

Having wrestled with the problems of subject and keyword searching over my entire 30-year career in industrial research, I have come to the conclusion that a solution to this dilemma lies in classifying technologies by the functions they perform.

[ . . . ]

Hence, functional classification of technologies ought to enable technical databases to be much more effective tools for problem solving by establishing a common language between people with problems and databases with answers.

Ernest J. Breton, "Functional Representation of Technology," in Advances in Classification Research, p. 25

Of course, knowledge management isn’t just a technology, and knowledge management lacks historical parallels to such categories as diagnosis and treatment in medicine. Nor is there a real-world artifact around which a systematic description of knowledge management can be formed — as the human body serves as the foundation for ontologies in medicine.

The problem is much bigger than even these difficulties imply, because knowledge management activities are pervasive and wildly heterogeneous, and knowledge management is often decidedly non-physical, reaching out to include such abstract topics as states of knowing and their impact on doing.

Why a classification of knowledge management functions is important

As great as the difficulty of creating a classification of knowledge management functions may be, the benefits are even greater. How important do we think this effort is? We predict without fear of contradiction that semantic classification of business functions and of other facets of the intellectual capital of organizations will, itself, become a core knowledge management activity that serves as

How to help us get started

To get the ball rolling, please fill out the form at the beginning of this page, or contact Phil Murray to express your interest.

Phil Murray
Editor-in-Chief
Knowledge Praxis
pmurray@media-access.com

Direct voice line: 757-397-4311


Top | Return To Knowledge Praxis | Knowledge Praxis Forum