public class BackwardConjunctionReductionRule extends AggregationRule
s1 and s2, this rule elides any
constituent in the right periphery of s1 which is
form-identical to a constituent with the same function in
s2, that is, the two constituents are essentially identical in
their final, realised, form.
The current implementation is loosely based on the algorithm in Harbusch and Kempen (2009), which is described here: http://aclweb.org/anthology-new/W/W09/W09-0624.pdf
Implementation note: The current implementation only applies ellipsis to phrasal constituents (i.e. not to their component lexical items).
*
Note:: this rule can be used in conjunction with the
ForwardConjunctionReductionRule in Aggregator.
factory| Constructor and Description |
|---|
BackwardConjunctionReductionRule()
Creates a new
BackwardConjunctionReduction. |
| Modifier and Type | Method and Description |
|---|---|
NLGElement |
apply(NLGElement previous,
NLGElement next)
Applies backward conjunction reduction to two NLGElements e1 and e2,
succeeding only if they are clauses (that is, e1.getCategory() ==
e2.getCategory ==
PhraseCategory.CLAUSE). |
apply, apply, getFactory, setFactorypublic BackwardConjunctionReductionRule()
BackwardConjunctionReduction.public NLGElement apply(NLGElement previous, NLGElement next)
PhraseCategory.CLAUSE).apply in class AggregationRuleprevious - the first phrasenext - the second phrasenull otherwiseCopyright © 2020. All Rights Reserved.