Technologies in the field of information retrieval, knowledge management, and communication technology currently face fundamental barriers, particularly in knowledge resource generation and machine language understanding, denying government and industry access to critical information resources. StreamSage proposes to solve these problems with its Property-Based Hierarchy Construction system, the first automatic ontology generator to successfully accomplish the following: extract, align, and organize all relevant space launch and range concepts into a Unified Missile Defense Ontology; associate relevant documents as well as contextually relevant sections of those documents with each concept in the ontology; and provide a bridge between structured and unstructured data by relating both to the searchable and browsable UMDO. StreamSage currently possesses state-of-the-art tools for several key component problems, which it will employ to overcome the three fundamental barriers to ontology generation: word sense ambiguity, term redundancy, and automatic learning of ontological relations, through a path-breaking combination of mutual-information, term-clustering, word sense disambiguation, and text-mining techniques. This project will not only overcome existing knowledge management barriers, but will enable future capabilities such as automatic domain updates, personalized information delivery, and automatic mission-crucial information discovery and extraction
