I think knowledge management tools are best categorized on a continuum, the salient points on which and representative programs being:

1.Recording essential decisions about a writing or product , organizing, and building on them, in other words, brainstorming and analysis. BrainStorm (www.brainstormsw.com), NoteMap (www.casesoft.com)
2.Organizing the parts of a complete writing. Lonely Notes, Writers Blocks, outline view in Microsoft Word.
3.Recording and organizing background material along with the material to be included in writing or directly relevant to a project. Miss Lonely Notes (www.mindola.com), Writers Blocks.
4.Creating a large and ever-expanding knowledge base, e.g. for all of oneís professional activities. Idea! Professional Edition, Personal KnowBase, generic database trees such as HyperClip.
5.Creating a knowledge base for a group or just a still larger knowledge base. Idea! Enterprise Edition, MDE InfoHandler , Zoot!

This dimension is one of increasingly broad focus of the product, starting with a core of central information and structural decisions about a project. Moving top to bottom, the focus is increasingly from transparency of operation and each of radical restructuring to searching and structuring to find. One implication of this scheme might be that redundancy increases when programs are close to one another on the continuum. By hypothesis it makes little sense to use BrainStorm and NoteMap for a single project. (I use NoteMap to format a BrainStorm model, but selecting a $150 program to assist a $70 program would not have been frugal.

Stephen R. Diamond

Where does ADM 3 fall in this scheme. By intent, it covers the full spectrum from 1 to 5, or at least 1 to 4. What compromises does its expansiveness impose? The outliner as such as good, but not as good as BrainStorm or even NoteMap. NoteMap, for example, gives you numerous ways to accomplish any operation; ADM provides fewer. To do otherwise would crowd the interface and reduce its usefulness.

Some would question the usefulness of powerful outlining features in a product that aims mainly mainly at level 4. But they are no doubt useful in a database, since you might discover and altogether better classification scheme for your data. There is also a definite efficiency in doing your brainstorming in the same application as your data collection, even versus an interoperable program. But are these efficiencies enough to overcome the loss of features in each category entailed in a high-concept application (as I think Alexander dubbed ADM)? Does it overcome the problems such universality poses, such as not paying enough attention during development to functions essential for one or another of (my) levels, resulting in not including a powerful undo functionality, which is essential to an outliner and is appropriately very powerful in both BrainStorm and NoteMap?

I can't answer this except by awaiting the further development of ADM. One basic design choice, however, decidedly emphasizes the tree database character of ADM, and that's the multiplication of views based primarily on the number of panels and their position. Most tree databases over two panels, but some offer three. ADM offers everything from 1 1/2 to 3. This refinement increases the usefulness of ADM as a tree database, but it doesn't preserve a proportion about the various levels of functionality (1 through 4 or 1 through 5). This isn't a ground level decision I would have made to increase the power of an outliner.

Stephen R. Diamond

Stephen,

I think I can buy into your taxonomy. Very interesting.

One use that you did not mention, perhaps intentionally, is the function of actually assembling the pieces you've been collecting and organizing for a specific project, and adding the cohesive text that turns it into a narrative... i.e. actually writing the piece. Most of us now probably export our pieces into a word processor for that stage because few, if any, of these programs are good at creating the final text (and I don't mean the formatting, which is another function altogether). To me, a "perfect" outliner would also be able to serve this function. Not to beat a dead horse, but I found that GrandView provided the tools for creating the entire work, from data collection to final composition.

The problem with the two pane outliners is that each piece of material is treated as a kernel or discrete piece of data, associated with the pieces around them, but not actually of those pieces. That's why I think a single pane outliner like GrandView provides the best writing environment. NoteMap doesn't deliver because it doesn't provide the cataloging/categorizing/identifying of those kernels the way ADM, Zoot, etc. do. That cataloging function is necessary, especially on big projects with lots of data, for helping the writer to manage all the information effectively. If NoteMap would add meta-data functions for each of the notes, then it would go a long way toward bridging that gap... assuming, of course, that one could sort, search and filter on the meta-data fields.

Would you agree with this, or do you have another perspective?

Again, thanks for providing the taxonomy.

Steve Z.

Stephen, thanks a million for your taxonomy. I realised that I have mainly been using tools aimed at 1 (brainstorming) and 4 (creating a generic knowledge base) as well as pushing them to cover needs 2 (structured writing) and 3 (organising background material).

My own way of thinking about your taxonomy is an office (material, not virtual). The brainstorming tool (1) might be a blank piece of paper on the desktop; next to it are one's writing notes (2); stacks of photocopies of the main reference texts and press clippings (3) lie on the desktop; books and magazines on the wider thematic spectrum (4) lie on the table and the floor; finally, the bookshelf and archive box are the larger knowledge base (5).

On a more horizontal approach, I would add the ability to link knowledge items to structured data, i.e. contacts, communications, actions and time/date. On the one hand this might concern the information itself, i.e. for a journalist or lawyer, the source of an item might be a person. In other cases this concerns work organisation, i.e. who should be on the recipient list for one's latest article. As with e-mails, I find that the current range of products offers very limited connectivity on this aspect.

In the office metaphor, a sticky note on the knowledge item is all that would be required.

alx

Stephen,

Back to your KM Taxonomy, where would you classify UltraRecall? To me, it seems to fit in category 4 (creating a large and ever-expanding knowledge base)

alx