Dynamic Data Mining

The goal of the project is to produce an innovative information software architecture for intelligent, dynamic data mining; one providing direct support for intelligent agents and scalable to very large data/knowledge bases. The objectives are to develop a new technology, deductive data mining, and use it to harness necessary extensions of inductive data mining technology needed to address video, fax, graphics, images/ voice, and textual data. The focus of the effort is providing advanced data structures that support innovative ways to both access and store various forms of data/knowledge. Specifically, the technology provides intelligent ways to coordinate various forms of raw data, including restructuring, to discover information. This is done by providing adaptable transformation techniques to structure knowledge so that it can be managed more efficiently, and so that information can be automatically filtered, manipulated, and summarized. The result is an evolvable data/knowledge base configuration for scalable information aggregation/ processing. The initial phase investigates development of techniques for designing, developing, and integrating large-scale active information systems using massive multi-source data rich repositories.