A94-1022 |
complete system including the
|
Semantic Linker
|
was evaluated in the December
|
A94-1022 |
experiments show that using the
|
Semantic Linker
|
reduced our system 's error rate
|
A94-1022 |
we discuss subsequently ) the
|
Semantic Linker
|
has been show to dramatically
|
H93-1007 |
LINKS AND THEIR PROBABILITIES The
|
Semantic Linker
|
first computes the link database
|
H93-1007 |
with an earlier w ~ rsion of the
|
Semantic Linker
|
than that reported here . In
|
A94-1022 |
a set of semantic graphs . The
|
Semantic Linker
|
seeks to connect these sub-graphs
|
H93-1007 |
parsing . This paper presents the
|
Semantic Linker
|
, the fallback component used
|
A94-1022 |
the relation AIRLINE-OF ) . The
|
semantic linker
|
may also " hallucinate " a new
|
A94-1022 |
very different algorithm , the
|
Semantic Linker
|
uses the same set of realization
|
A94-1022 |
the Delphi system is called the
|
Semantic Linker
|
( Stallard and Bo - brow , 1993
|
H93-1007 |
features influence the actions of the
|
Semantic Linker
|
with respect to such matters
|
A94-1022 |
produced . Links established by the
|
semantic linker
|
are treated by the quantification
|
H93-1007 |
ABSTRACT This paper presents the
|
Semantic Linker
|
, the fallbaek component used
|
H93-1007 |
The important new feature of the
|
Semantic Linker
|
's topic choice module is its
|
A94-1022 |
runs over them again with the
|
Semantic Linker
|
enabled . 8 Results Of Formal
|
H93-1007 |
evaluation -LSB- 6 -RSB- show that the
|
Semantic Linker
|
reduced DELPHI 's Weighted Error
|
A94-1022 |
of the utterance . Because the
|
Semantic Linker
|
uses the same set of realization
|
A94-1022 |
other syntactic ev - idence . The
|
Semantic Linker
|
does not have these re - quirements
|
H93-1007 |
spoken language system HARC . The
|
Semantic Linker
|
is invoked when DELPHI 's regular
|
H93-1007 |
spoken language system HARC . The
|
Semantic Linker
|
is invoked when DELPHI 's regular
|