regelcomputer

Regelcomputer, or rule computer, refers to a class of computing systems that derive results by applying a predefined set of explicit rules to input data. Unlike conventional procedural programs, regelcomputers rely on a rule base and an inference mechanism to deduce conclusions or outputs from facts.

An regelcomputer’s architecture typically consists of three parts: a rule base containing production rules in the form if-then, a working memory holding facts about the current situation, and an inference engine that matches rules to facts and fires applicable rules. The execution can be data-driven using forward chaining, where new facts are inferred from existing ones, or goal-driven using backward chaining to determine which rules must fire to achieve a desired conclusion.

Historically, regelcomputers are associated with production systems and expert systems from the 1970s to the 1990s. They influenced modern business rules management systems and various rule engines in software development. They share kinship with logic programming and rule-based AI, but they are distinct from purely procedural algorithms or statistical learning models.

Common implementations include production systems such as OPS5 and CLIPS, and contemporary equivalents appear in business rule management systems and workflow automation tools. They are favored in domains requiring transparent decision logic, explainability, and the ability to modify rules without changing underlying code.

Limitations include potential rule conflicts, maintenance complexity as the rule set expands, and performance concerns with very large knowledge bases. Regelcomputers often complement other programming paradigms rather than replace them, and are frequently used in hybrid systems that integrate rule-based reasoning with data-driven components.