Definition
An expert system applies an explicitly represented knowledge base and inference rules to make recommendations or decisions within a bounded domain. Classic systems separate domain facts from an inference engine that evaluates rules such as if-then conditions.
Distinguish it from nearby terms
Expert systems encode much of their decision logic explicitly; machine-learning models learn statistical parameters from data. Modern systems can combine both, using rules or policy engines to constrain probabilistic models.
Check your understanding
Explicit rules can improve traceability but still fail when the knowledge base is incomplete, stale, contradictory, or applied outside its intended domain.