Recently Completed Programs Home
Overview
Mission
Background
Vision
Technical Program
ACIP Phase I
Approach
Challenges
Goals
|
|
|
|
Architectures for Cognitive Information Processing (ACIP)
Background:
Military forces today and increasingly for the future must be able to quickly react and adapt to complex and dynamic missions and scenarios. The complexity of the battlefield and the collage of information sources that must be brought to bear commonly overwhelm battlefield systems and the ability of the warfighter to assimilate and comprehend important information. Embedded, real-time "cognitive" processing for both the warfighter and associated automated systems will be critical to success in this complex environment (i.e., algorithms that do logical probabilistic reasoning over large knowledge bases, and also use learning techniques to improve over time). Current intelligent processing implementations depend on the use of existing COTS computing architectures that were developed and are best suited for numeric processing applications. In addition, today's knowledge representations, abstraction (processing objects), architectures, and implementations are ad hoc, awkward and inefficient. To realize the impact and promise of cognitive information processing approaches, computing architectures and development frameworks attuned to cognitive processing fundamentals need to be researched, identified, and developed to implement uniquely cognitive processing structures and techniques efficiently.
|
