Inexpensive Intelligence Using Procedural Propositional Logic

Abstract

Implementations of inference engine systems invoke many costs, including the cost of the inference engine itself, the cost of integrating the inference engine, and the cost of specialized personnel needed to create and maintain the system. These costs make a very high return on investment a criterion for incorporating these systems into the corporate portfolio of applications and technologies. Recently, the No Inference Engine Theory (NIET) [8] has been developed for creating procedural propositional logic rule-based systems. The NIET systems are implemented in traditional procedural languages such as C++ and do not need an inference engine or proprietary languages, thus eliminating the cost of the inference engine, the cost of integrating the system, and the cost for knowledge of a proprietary language. In addition, these procedural systems are an order of magnitude faster [8] than inference systems and maintain linear performance. For problems using propositional logic, the procedural systems described in this paper offer dramatically lower costs, higher performance, and ease of integration. Lowering the external costs and eliminating the need for specialized skills should make NIET systems more profitable and lead to the wider use of propositional logic systems in business.