Omnity is a self-assembling, automated knowledge curation and discovery  system enabling enterprise knowledge workers to cope with the explosion of knowledge all around them.  Omnity finds shared meaning between documents by detecting similar patterns of shared rare words (semantic signatures) – a mathematical approach that transcends the limitations of link-based document discovery.  Representing its’ core meaning, the semantic signature of a document is formed both from the static, publication date-fixed content of the rare words in that document compared to the time-dependent evolving context of related results around that document. Omnity forms an updatable semantic signature for each document based on (a) the mathematical signature of the statistical distribution of the rare words in a document by use of natural language processing, (b) the evolving “connective shape” of the sub-graph of knowledge in which a query document attracts resulting documents within an insight graph, leveraging graph mathematics, and (c) the semantic and graphical comparison of the query and response documents, by use of machine learning.