P00-1034 |
simplified version of the well-known
|
back-off smoothing
|
method is used . To mitigate
|
P00-1034 |
simplified version of the well-known
|
back-off smoothing
|
method is used . To mitigate
|
J14-4003 |
Factored Language Models with
|
back-off smoothing
|
. Section 5 presents two methods
|
E97-1032 |
. Employing a small amount of
|
back-off smoothing
|
also for the known words is useful
|
J15-2001 |
suffer from data sparsity and the
|
back-off smoothing
|
may fall back to very short contexts
|
D13-1024 |
sequence . Models based on so-called
|
back-off smoothing
|
have shown good predictive power
|
C00-1070 |
lexicalized models use a simplified
|
back-off smoothing
|
technique to overcome data Sl
|
J04-2004 |
deal with unseen n-grams , the
|
back-off smoothing
|
technique from the CMU Statistical
|
C96-2151 |
idea is usually referred to as
|
back-off smoothing
|
, see ( Katz 1987 ) . These techniques
|
J14-4003 |
using Witten -- Bell interpolated
|
back-off smoothing
|
, according to the back-off graphs
|
J02-3004 |
no increase in performance for
|
back-off smoothing
|
( see Tables 7 and 8 ) . These
|
C00-1070 |
the formalism of our simplified
|
back-off smoothing
|
, each of probabilities whose
|
J02-3004 |
necessary for class-based and
|
back-off smoothing
|
, we maintain the train/test
|
C00-1070 |
using the perplexity measure , a
|
back-off smoothing
|
( Katz , 1987 ) is said to perform
|
J02-3004 |
1999 ) use perplexity to compare
|
back-off smoothing
|
against distance-weighted averaging
|
N13-1117 |
processing due to their relationship to
|
back-off smoothing
|
. Denoting a context of atoms
|
C94-1023 |
respectively . After applying
|
back-off smoothing
|
( Katz 1987 ) and robust learning
|
D15-1231 |
This can be viewed as a kind of
|
back-off smoothing
|
( Katz , 1987 ) , where the Observed
|
J02-3004 |
organization of conceptual information .
|
Back-off smoothing
|
, however , incorporates no notion
|
C96-2151 |
' k - ~ , Tk-1 ) . 3.1 N-gram
|
Back-off Smoothing
|
We will first consider estimating
|