ABSTRACT
Extended network expert operating system (ENEOS) models the thought processes of a team of human experts when it assesses the network environment and determines the most reasonable solution to any problem, choosing from a pool of existing possibilities. In ENEOS, the knowledge bases and the associated inference engine of the finite inference frame processor have separate components for encoding control knowledge, factual knowledge, and judgmental rules. To provide a transparent presentation of both control knowledge and factual knowledge, the knowledge bases in ENEOS are organized in distinct frames that contain an ordered rule structure of separately encoded control blocks. Control knowledge, like rule-clause ordering, specifies when and how a program should perform such operations as pursuing a goal, acquiring data, focusing on an object, and making an inference. Correlating events for problem determination and analysis is not a straightforward task; it is affected by time correlation, event attenuation, redundant events, location correlation, partial data, and multiple problems.
