The Acquisition of Procedural Knowledge through Inductive Learning
Most induction systems previously developed are concerned with the acquisition of generalized characteristic descriptions from training instances represented in either attribute tuples or first order predicate (or their variations). An extension to the problem can be achieved by allowing each training instance to take the form of a sequence of descriptions, which may represent an instance of an operation sequence, or the description of a temporal event.
KeywordsProcedural Knowledge Training Instance Induction System Canonical Representation Operation Sequence
Unable to display preview. Download preview PDF.