KASER: A Qualitatively Fuzzy Object-Oriented Expert System Inference Engine

摘要

This paper details the results of two years and one million dollars invested by the ONR towards the development of a qualitatively fuzzy expert system shell. The shell has been developed for the express purpose of fuzzy qualitative reasoning. The relation among object predicates is defined by object trees, which are fully capable of dynamic growth and maintenance. A result of the development of a qualitatively fuzzy inference engine is that the expert system can then acquire a virtual-rule space that is exponentially (i.e., subject to machine implementation constants) larger than the actual, declared-rule space and with a decreasing non-zero likelihood of error. This capability is termed, knowledge amplification and the methodology by which it may be achieved is termed a KASER. KASER is an acronym for Knowledge Amplification by Structured Expert Randomization. It can crack the knowledge-acquisition bottleneck in expert systems. The KASER represents an intelligent, creative system that fails softly, learns over a network, and has enormous potential for automated decision making. KASERs compute with words and phrases and present capabilities for metaphorical explanations. The software will be demonstrated at the conference proper.