This simplification and division of data is very similar to a classification hierarchy, except that it doesn't relate data hierarchically to only one or two parent nodes in a classification tree. Instead, adaptive clustering relates data to any number of other nodes in and across its immediate classification tree. These nodes are experientially associated, given previous data in their memory banks. As a result, adaptive clustering, through its method of divide and conquer to maintain simplicity, grows multiple classification hierarchies in real-time, on-the-fly, which are interrelated in multiple dimensions.

Hyperlinked classification taxonomies facilitate discovery
Classification nodes appear at varied levels in the system of multiple classification trees as they reach different levels of significance. Since significance can be determined at several levels in the hierarchy in a relative fashion, adaptive classification nodes effectively self-promote themselves up and down a classification tree according to their strength. This creates a rich dynamic knowledge base of interrelated classification nodes, which point to the paragraphs in the source documents from which they came. This dynamic, interrelated classification tree database provides a rich multi-taxonomy publishing and information discovery environment.

Unlike neural networks, which represent static, one dimensional, structured classification models simulating a snapshot of the brain, adaptive classification taxonomies actually emulate how the brain works, continuously and dynamically growing multiple classification hierarchies in multiple dimensions. These multiple hierarchies represent the current reality of the entire domain of data at their disposal. Like baking a cake, data can be added to increase the flavor and change the texture of the resulting interrelated classification hierarchies.

Adaptive classification hierarchies provide significant added value
Adaptive classification hierarchies deliver a lot more value beyond the domain classification task. With Adaptive processing, each document that's created or changed is processed immediately with little more time spent than is required to save the document to disk. This creates a document content signature that represents a mini-classification hierarchy of the document's content. Each node receives a value representing its strength within the context of the document.

Document classification hierarchy nodes can easily be navigated for discovery, and aggregated to provide a network of interrelated hierarchies effectively representing a latticework, dynamically emulating the way the brain works. Classification nodes can be massed together and published through tools like XML (extensible markup language) in a browser or Lotus Notes view to represent multiple classification taxonomies. This is much like multiple subject views of a database provided by multidimensional OLAP (online analytical processing) tools, based on the significance of aggregate node values.