Definition
Systems that organize domain knowledge and reasoning processes to provide analysis or advice within a limited problem area.
Early DENDRAL used chemical knowledge to constrain the search for candidate structures, demonstrating the importance of domain knowledge in problem-solving. This entry covers reusable ways of organizing knowledge and reasoning, with DENDRAL serving only as a historical example.
Expert systems generally require ongoing maintenance of their knowledge and conditions of applicability. Performance within one domain cannot establish understanding across domains, nor does it mean that professional responsibility can be handed over to a program.
Principles and impact
Problems solved
- Organized experience and reasoning methods from a specific domain into reusable programs.
Principles
- A knowledge base works with a reasoning process, using domain constraints to narrow the set of candidate solutions.
Improvements
Not recorded
Significance
- Helped move AI from exploration of general problem-solving into specialized applications.
History and applications
People
Edward Feigenbaum、Bruce Buchanan、Joshua Lederberg
Organizations
Not recorded
Events
Not recorded
Applications
Screening scientific hypotheses、Domain-specific decision support
Limitations
- Knowledge is difficult to acquire and costly to maintain, and systems can easily fail outside the scope of their knowledge.
Upstream and downstream relations
Upstream
Sources: DENDRAL — Original Papers and Software ArchiveComputer Science as Empirical Inquiry: Symbols and Search
Downstream
No direct downstream relations recorded