|KM Archive Home Site Guide I3Update Insights Resources K-Shop About Us Contact Us ::Our new (2011) KM website is kmknowledge.com ::|
September 2002 Feature
a free monthly briefing on the knowledge agenda
The A-Z of Knowledge Technology (Part 1)
Any follower of knowledge management will know that the things that change most (apart from your client or bosses minds!) is that of KM technology. Five years ago, Lotus Notes was probably the most cited knowledge sharing technology among KM leaders, while products like content management suites or expertise profilers barely existed. It's a job to track and keep on top of significant developments, yet alone predict how thing will unfold. Even the IT analysts frequently get it wrong. Therefore, based on a recent request to give a KM technical briefing, I have taken a somewhat different approach to analyzing technological developments.
The approach is an A-Z, identifying developments of note for each letter of the alphabet. Be warned; this is not the most logical framework to position developments (there are many Cs and Ps, so selecting which one to include is my personal choice; on the other hand things get sparse at the back end of the alphabet), but it is interesting and should trigger some of your own thoughts (share them with us in the next I3 UPDATE / ENTOVATION International News). In this edition we cover the first half of the alphabet.
A: Artificial Intelligence. Remember in the 1970s when AI was going to be the pervading knowledge technology? Things did not happen like that, because AI lends itself only to specific types of problem - it struggles in dealing with "common sense". However, developments in areas like natural language processing, neural networks, genetic algorithms (even ant algorithms) mean that it is holding a definite niche in knowledge discovery such that insurers can more accurately asses risk, or retailers identify clusters of products that are bought together.
B: Bayes and Boole. Not the name of a travel agent, but of two people who set some of the foundations for modern search engines. The Reverend Thomas Bayes (c. 1702-1761) developed a treatise on probability. Search engines like Autonomy use a Baysean approach (frequency of word use, clustering etc.) to ranking the relevance of documents, and because no natural language analysis is involved, it can deal easily with multiple languages. George Boole (1815-1864) is known for Boolean algebra (AND, OR, NOT). But how many of you actually get beyond the front page of Google to its "advanced search" where you can filter your search results by applying the underlying Boolean constructs?
C. Content Management. These systems (CMS) have revolutionized the development of websites and intranets over the last few years. The basic notion is one of entering once, and using it many places. Information is granularized into small reusable chunks. A single Web page may be composed of many individual content items that can be personalized to the users' needs. There is growing convergence between what were once separate solutions of document management and information portals. In all of these good metadata to describe the components and their uses is important, so expect semi-automatic classification of content (see Retrieval) to assume a greater role.
D. Document Management. Since many electronic document management solutions (EDMS) have embraced the features of content management systems, where does that leave document management? Well, documents are still the dominant way that many professionals evolve and publish their knowledge. A good document management system will also manage many legacy documents, files in different formats and those scanned in. Overall, though, the main trend is towards segregation of function, so that different components of EDMS and CMS can inter-operate in a broader portal environment.
E. Email. For all the developments in CMS etc., email is probably still the dominant means of knowledge sharing within and beyond organizations. More email was sent in 1999 that all the previous yearx combined. The main problem today is coping with this email deluge, much of which might not be relevant to the user's immediate needs. AI comes to the rescue in terms of filtering agents and natural language analysis that identifies experts and categorizes key concepts for later retrieval, as in systems like AskMe and Tacit's Knowledgemail.
F. Framework. A framework provide positioning of all the technology options relative to each other. One that I find useful has three dimensions 1) phase of the knowledge life cycle: create, discover, gather, classify, store, retrieve, use; 2) level of use: individual, team, organization, inter-organization; 3) type: highly structured (as in database), semi-structured (e.g. metadata plus free text), unstructured (free text), tacit (in people). Some technical solutions apply only to a given cell or row. Others, notably enterprise solutions or 'suites' cover a broad spectrum. Activity within the framework is like an ecology, with some cells growing in importance, others withering; with some vendors extending their scope and various rationalization taking place in adjacent cells (c.f. Inktomi's recent acquisition of Quiver - search engine adds classification, and Microsoft's acquisition of XDegrees P2P software)
G. Groupware. A name that is going out of fashion, but the problem is that there is a confusion of other names e.g. collaboration software, bulletin boards, computer conferencing etc. that often mean different things to different people. Essentially groupware covers a range of collaborative software solutions that span time and space. Some of the synchronous tools e.g. instant messaging, chat are becoming an important part of the enterprise tool set. Groupware is often an add-on to portal and other software, However, some organizations prefer to have dedicated and low cost solutions that are optimized for specific purposes c.f. Shell's use of Sitescape for online communities, or eRoom (project rooms) used by companies like Ford and Aventis.
H. Humans. Perhaps one the best technologies. Humans are generally smart and have common sense. Most of an organization's knowledge is tacit knowledge. Therefore, perhaps when considering KM technologies, we should spend more time looking at technologies that help humans share and evolve their knowledge better, rather than focusing on those that capture human knowledge into databases.
I. Intelligent Agents. The reverse initials of AI, but using AI techniques, typically to scour the Web and find relevant information and to alert users of updates. Many solutions that started as search engines (e.g. Verity and Autonomy) are now broader in scope and have intelligent agents that assess what a user is working on and alert the user to links to relevant content - a forerunner of truly adaptive systems that infers what a user is working on and offers appropriate assistance (but can you put up with the interruptions?).
J. JIT (Just In Time) Knowledge. Remember 'push' technology? Well, in one form its still hanging in there (just!). One of the poisoners of 'ticker tape' feeds for PCs - BackWeb - now offers 'polite' communications and flash alerts. Another form - email - pushes at you every day (even every minute of every day), but can you find that relevant email afterwards? Users want the relevant knowledge at the time they do a task. Therefore, perhaps the best approach to JIT is really excellent Retrieval (q.v.). See a reader's comment on this definition.
K. Know-bot (knowledge robot). More online shopping sites now have online 'assistants' to help you with your selection and purchase (see eGain's "assistant". These use techniques like case-based reasoning to assess your situation and offer suggestions, giving you also the option to talk to a real human as well. Yoda's helpdesk at Lucas Arts is another example. So why not extend this idea to in-house help-desks and knowledge centres?
L. Learning Objects. This is another case where increasing granularity (modularization) is boosting efficiency. With the growth of elearning, it makes sense to package the knowledge needed by the learner into relevant chunks. These can operate several levels, from a learning object addressing a small item of leaning taking just a minute or so, to a full lesson that combines serval other objects in a particular combination. Companies like Cisco and Dow are reported to have saved tens of millions of dollars by reorganizing their leaning material this way.
M. Mobile KM. As mobile phones and PDAs get smarter and access greater bandwidth, the scope for mobile knowledge management increases. More executive on the move will want access, not just to email, but corporate and external information resources that can be downloaded on the move and read it on a small screen (another reason for organizing core information repositories into small sized chunks). Combined with location specific knowledge, a whole new angle KM is emerging.
The KM technology alphabet concludes in the next issue. We welcome your feedback and own alphabetical favourites.
Top of Page
Home | Search | I3 UPDATE | Insights | Resources | About Us | E-mail Us
Copyright. All Rights Reserved.
Legal & Privacy Notice.