1. ProblemDefinition – Define the scope of the problem to be solved.
2. IdentifyDomainExpert - Identify who is the domain expert.
3. DataAcquisition – Acquire data that represents instances of the problem to be solved.
4. KnowledgeEngineeringProcess? – Acquire knowledge from the human expert.
5. DataRepresentation – Define the representation of data in the expert system.
6. RuleImplementation – Implement rules for the knowledge acquired in the KnowledgeEngineeringProcess? step.
7. RuleDebugging? – Debug the rules implemented in the RuleImplementation step.
8. TestingStrategies – Test the rules using a large population of problem instances.
9. IterationCycle - Go to the DataAcquisition step, and repeat until the expert system correctly solves a satisfactory percentage of problem instances.

Submitted by:
GeorgeWilliamson
Union Pacific Railroad
gawillia@up.com