Rule-based system

A computational system whose behavior is specified through explicit rules written by people. It applies those rules to facts or inputs to derive conclusions or actions.

Production system: A rule-based architecture that uses condition–action rules, a working memory, and a control strategy. Its rule-firing cycle is a canonical way to execute explicit rules.
Production rule: A conditional rule that specifies an action or conclusion when its stated conditions hold. It is the basic executable unit in many rule-based systems.
Expert system: A program that uses encoded specialist knowledge to solve problems in a specific domain. Expert systems are a prominent application of explicit knowledge and inference rules.
Machine learning: Methods that enable computer systems to improve performance by learning patterns from data. Learned models derive behavior from examples rather than relying solely on hand-written rules.
Inference engine: Software that applies logical rules to facts to derive conclusions or select actions. It determines which rules apply and how their consequences enter the system.
Knowledge base: An organized collection of facts, rules, and other information used by a computational system. It stores the domain knowledge that the system applies.
MYCIN: An early medical expert system that recommended treatments for bacterial infections using rules and certainty factors. It demonstrates how hand-authored rules can support a specialized diagnostic task.
Decision tree: A model that represents decisions as tests arranged in a branching tree. Decision trees encode decision logic in a structured form that can be learned or authored.
Forward chaining: An inference strategy that repeatedly applies rules whose premises match known facts. It lets rule systems derive consequences outward from available facts.
Boolean logic: A formal system for reasoning with statements that are true or false. Rule conditions often use Boolean operators to combine tests.

Linked from 10 pages