Semantic routers reflect the way
Information can route itself dynamically, without user intervention through a Knowledge Management tool called a semantic router. Profiles can be stacked in semantic routers, providing a set of conditional routing patterns that work in a way similar to a prism, through a method called semantic refraction. Semantic refraction derives a routing coefficient based on the goodness-of-fit of kSigs and profiles to router profiles. These then get submitted to a routing policy document, somewhat like a routing table, for conditional routing resolution. Routing policies can be as simple as Threshold levels--30% = alert, 50% = email, 70% = pager, 90% = workflow (node, condition)--or comprehensive routing rules as are deemed appropriate and consistent with corporate knowledge sharing and distribution policies needs and objectives.

K…K…who's there?
Semantic routers have other powerful uses. For example, rather than use them to route documents, they double nicely as a people finder. By stacking user profiles in semantic routers, you can run documents and profiles through the router and configure the routing policy to transform the routing slip into a people finder. So, if you wanted to find experts who have certain characteristics or knowledge, then you merely enter or copy and paste a representative text snippet or submit a representative document or profile, and the semantic router will return those users who fall within the resultant hyperspatial semantic containment.

I'm a K you're a K
The knowledge we apply to solve corporate problems and leverage corporate opportunities is reflected in the documents and email we write, as represented through corporate knowledge transactions. This is key to the knowledge sharing stage of the corporate knowledge management life cycle, since we learn new knowledge based on what we already know. Building knowledge in this way reflects our need to learn incrementally, thereby supporting the incremental learning paradigm, a fundamental principal of 2G-KM.

Because of the dynamically adaptive, high-dimensional incremental indexing capability in 2G-KM, documents can be combined dynamically and responsively by their common concepts into Knowledge ViewPoints as expressed through aggregate kSigs in a fashion that facilitates recursive semantic propagation, forming an aggregate knowledge concept tree. The resultant knowledge concept trees form the basis of a dynamically adaptive intelligent inferencing engine that supports not only causal reasoning through induction (forward chaining) and deduction (backward chaining), but also through abductive inferencing (the basis of machine learning) as represented by an optimal outcome tree from any node in a knowledge base latticework.

Lovin' to learn
To leverage our natural incremental learning strategy, 2G-KM technology must leverage how we are built. In order to better understand the power of this statement, permit me to take you into the depths of how we process information in our brains, so that you have a workable model as an image on which to build your knowledge about 2G-KM.