Catalogue of Artificial Intelligence Techniques
Negation as Failure
Keywords: Micro-PLANNER, closed world assumption, rule of inference
Categories: Logic Programming
Author(s): Martin Merry
Negation as failure is a rule of inference which allows one to deduce that NOT P is true if all possible proofs of P fail. This is the way that negation is treated in Prolog and Micro-PLANNER. When using a system for database access, the assumption that negation as failure corresponds to true negation is precisely a consequence of the closed world assumption, i.e., the assumption that all relevant information is contained in the database. Without the closed world assumption, negation as failure corresponds to `we assume something is false if we cannot deduce it from available information', which is not the same as true negation. Negation as failure is an example of Default Reasoning. Its mathematical meaning is extremely questionable.
- Reiter, R., On closed world databases, Readings in AI and Databases (Mylopolous, J. and Brodie, M.
, eds.), Morgan Kaufmann, San Mateo, CA, 1989, pp.248--258, also appears in Logic and Databases
(Gallaire, G. and Minker, J., eds.) Plenum Press,
New York and London, 1978, pp. 55--76