... spontaneous speech data and
present an analysis of the repair phenomena ob-
served. In addition, we describe ways in which
pattern matching, syntactic and semantic analy-
sis, and acoustic analysis ... some-
what planned (subjects pressed a button to begin
speaking to the system) and the transcribers who
produced lexical transcriptions of the sessions were
instructed to mark w...
... Shriberg and Stolcke (2001)
and include maximum F0 and energy of the spurt,
mean F0 and energy of the spurt, pitch contour (i.e.,
slope) and energy at multiple points (e.g., the first
and last 100 and ... PK WD
TOP 0.66 0.11 0.17
SUB 0.59 0.23 0.28
Table 1: Intercoder agreement of annotations at the
top-level (TOP) and sub-level (SUB) segments.
the annotators marked 8.7 top-level segments an...
... linguistic knowledge, general world
knowledge, and application domain knowledge,
DATALOG achieves a high degree of portability and
extendability.
1. Introduction
An area of continuing interest and ...
information from other, more general capabilities,
and thus have the ability to be transported from
one application to another. However, it is impor-
tant to realize that the d...
... verbs.
For each word, training and test instances tagged
with WordNet senses are provided. There are an av-
erage of 7.8 senses per target word type. On average
109 training instances per target word are ... Sun Yuan Kung.
Principal Component Neural Networks. Wiley,
New York, 1996.
Bradley Efron and Robert J. Tibshirani. An Intro-
duction to the Bootstrap. Chapman and Hall,
1993.
H...
... training data to a WSD
system. To reduce the effort required to adapt a
WSD system to a new domain, we employ an ac-
tive learning strategy (Lewis and Gale, 1994) to se-
lect examples to annotate ... estimation, and successfully used it
for probabilistic context-free grammar domain adap-
tation (Roark and Bacchiani, 2003) and language
model adaptation (Bacchiani and Roark, 2003)...
... closest to the target
word, and sends these words (and the target) on to
the next module for disambiguation. Note that these
words must all be known to WordNet, and should
not include any stop–words.
However, ... API to the Gtk toolkit. Unlike the command
line interface, the graphical interface is not tied to
any input file format. The interface allows the user to
input text, and...
... rather than KL distance in order to test
statistical significance.
L1 words that translate into the same L2 word
are grouped into clusters;
SALAAM identifies the appropriate senses for
the words in ... for sense annota-
tion. The key intuition behind SALAAM is that when
words in one language, L1, are translated into the
same word in a second language, L2, then those
L1 words are semantica...
... utterances with 13,190 words. These utter-
ances were hand coded for pitch accent and intona-
tional phrase brakes.
3.1 Pitch Accent Coding
The utterances were hand labeled for accents and
boundaries ... seem to play
a role in accentuation as well. The first word of in-
tonational phrases (IP) is less likely to be accented
while the last word of an IP tends be accented. In
short, acc...
... perceptual and acoustic
properties and syntactic and statistical determinants.
Speech and Language, 7, 333-371.
Umeda, N. and R. Teranishi. The parsing program for
automatic text -to- speech ... Letter -to- sound rules for automatic translation
of English text to phonetics. IEEE Transactions on
Acoustics, Speech, and Signal Processing, 6, 446-459.
Gee, J. P. and F. Grosjean. 1983. Pe...