Knowledge Management Manages Information Relations

The Dekstrus DNE knowledge management and object modeling tool is designed to manage and map information relations. In real life, pieces of information can be related in many different ways. While most knowledge management solutions allow you to relate them in only two ways, the DNE provides four strategies that you can mix and match, allowing you the flexibility the map any relation you believe holds between the information you are using.

The first and most common way to relate information is vertically, in a tree structure. Here information is managed in a one-to-many structure, where sub-topics are ranged under topics and pieces of information are ranged under sub-topics. This is a taxonomic division of information: complexity here only goes in one direction: any topic can have many sub-topics, but a sub-topic can only be found under one topic. Although nearly all digital knowledge management is done using a tree structure, tree structures are best suited for analog information. Think of an bookshelf as primitive KM. When organizing a bookshelf, you have several books, each of which can be put on exactly one shelf, so it is crucial create a structure so that each book ends up in the correct place so you can find it later.

Tree structures can be very useful for some purposes. However, when managing information digitally, we don't need to limit ourselves to a single strategy. Unlike with physical pieces of information like books on a bookshelf, with digital information we don't need to find the "correct" place for anything: we can create many links to the same information, organizing overlapping sets of information pieces in different ways. Because we are managing shortcuts to files rather than the files themselves, we have much greater freedom in how we document information relations. For instance, we can employ a KM strategy that allows us to relate any piece of information to any other:

This "affinity" map organizes information without the one-to-many limitation in taxonomies. Information relations mapped in this way include ontologies, topic maps and data models. Any piece of information can be related to any other, and you can see at a glance the various relations (both direct and indirect "relationship paths" through mutual affinities) that hold between different pieces of information. This captures more complex relations than is possible with a tree structure. Affinity maps give you the ability to map multiple sets of relations between pieces of information, rather than forcing each piece into a single "right" relation rather than any other.

However, while affinity maps allow you to document overlapping relations between many pieces of information, they do not indicate how the relation works (a tree structure, on the other hand, does indicate the relation between pieces of information - the relation is always a contribution relation from one to many - e.g., folder to sub-folder). A third way of relating pieces of information, a contribution relation, combines the flexibility of the many-to-many relations possible in an affinity with the directionality of relations possible in a taxonomy.

Here any piece of information can be related to any other (as in an affinity map), but with the added feature of directionality (as in a tree structure). The arrows indicate the direction of a contribution. Many object models, such as results chains and process models, are examples of this sort of relation.

The fourth sort of information relation is an intersection. An intersection between two collections of information will provide you with all the pieces of information that belong to either one or both (or all three, etc.) of those collections.

Here the middle three pieces of information belong to one of the collections or the other, and the middle piece alone belongs to one collection and the other. Search engines work by gathering together pieces of information that stand at the intersection of other groups of information.

As you can see, all four methods of relating information are very important, and a KM application ought to be able to flexible enough to allow you to organize your information in any of these ways. However, most KM applications do not have this power and flexibility. Most knowledge management applications consist of just a tree structure and a search engine.

The DNE uses all four information management strategies to provide you with the most powerful, most integrated and most intuitive KM solution.