Inferring negative information from disjunctive databases |
| |
Authors: | Kenneth A. Ross Rodney W. Topor |
| |
Affiliation: | (1) Department of Computer Science, University of Melbourne, 3052 Parkville, Victoria, Australia;(2) Present address: Department of Computer Science, Stanford University, 94305 Stanford, CA, U.S.A. |
| |
Abstract: | We propose criteria that any rule for inferring negative information from disjunctive databases should satisfy, and examine existing rules from this viewpoint. We then present a new inference rule, the ‘disjunctive database rule’ (DDR), and compare it to the existing rules with respect to the criteria. In particular, the DDR is equivalent to the CWA for definite databases, it infers no more negative information than the GCWA, and it interprets disjunction inclusively rather than exclusively. We generalize the DDR to a class of layered databases, describe an implementation of the DDR, ‘negation as positive failure’, and study its soundness and completeness properties. |
| |
Keywords: | Deductive database disjunctive database indefinite database inference negation closed world assumption |
本文献已被 SpringerLink 等数据库收录! |
|