... Saarbr
¨
ucken
{neumann|schmeier}@dfki.de
Abstract
We present a mobile touchable application for
online topic graph extraction and exploration
of web content. The system has been imple-
mented for operation on an iPad. The topic
graph is constructed ... Proceedings of the ACL-HLT 2011 System Demonstrations, pages 20–25,
Portland, Oregon, USA, 21 June 2011.
c
2...
... & Information Science
3
Dpt. Computer Sciences
Swedish Inst. of Computer Science Norwegian Univ. of Science and Technology Univ. of Tampere
Kista, Sweden Trondheim, Norway Tampere, Finland
{olovs,gamback}@sics.se ... to themobile part
and in return gets the results of the performed ex-
ercises from the mobile component.
65
Figure 2: Home Companion interface
2.1 The Home H&am...
... experiment 2 for
hexoses (black) and sucrose (grey).
Doc. S1a. Model structure of C24.
Doc. S1b. sbml format of model structure of C24.
Doc. S2a. Model structure of Rsch.
Doc. S2b. sbml format of model ... control of each assay was
boiled for 3 min after addition of enzyme extract. Reac-
tions were incubated for 60 min at 30 and 4 °C, and
stopped by boiling for 3 m...
... two
atoms (i,j), of types t
i
and t
j
, and N
obs
(d
ij
,t
i
,t
j
) represents the
number of atoms of types t
i
and t
j
observed in the structure
training set, separated by a distance of at least ... set of protein–RNA decoys for the RNA, DNA
and reduced atom distance potentials in decoy discrim-
ination.
Table S5. Comparison of the performance of the
potential with di...
... of the
Deutsche Forschungsgemeinschaft (NI 592/2-1). We thank Heinrich
Betz for his support and for critical reading of the manuscript.
References
1. Langley, J.N. (1907) On the contraction of ... the identification and
characterization of a-conotoxins with selectivity for neur-
onal nAChR subtypes and provides examples of their use in
defining the compositions and funct...
... and (Blei et al., 2003a)
for details and demonstrations of topic models.
816
Summary hLDA (sumHLDA): The hLDA
represents distribution of topics in sentences by
organizing topics into a tree of ... minimum number of topics (# of top-
ics=depth), while setting γ
o
= 10 creates exces-
sive amount of topics. γ
0
= 1 gives reasonable
number of topics, thus we use this value...
... model performs rather well for
a limited number of candidate structures, and out-
performs Charniak’s model when k = 5. In this
case we observe a small boost in performance for
the detection of junction ... prob-
ability of the k-best candidates using a standard
vanilla-PCFG model derived from the same train-
ing corpus. Moreover, we evaluate, by means of an
oracle, the upper a...
... Experiments and Analyses
4.1 Corpus Information
Table 1: Corpus information
Table 1 shows the information of corpus which is
used for experiments. All corpora are obtained
from Korean web chatting ... sentence
T
and candidates consist of
Eumjeol and the blank character .
C
i
s
i
b
nn
bsbsbsbsT
332211
=
.
332211 nn
bsbsbsbsC
=
(2)
(n is the number of Eumjeols)...
... n-character slice for text categorization by lan-
guage (Cavnar and Trenkle, 1994) and Phone Rec-
ognition followed by n-gram Language Modeling,
or PRLM (Zissman, 1996) .
Orthographic forms of language, ... to the
bag -of- words
paradigm originally formulated in
the context of information retrieval (IR) and text
categorization (TC) (Salton 1971; Berry
et al.
,
1995;
Chu-C...
... from the block set itself.
and are single-word blocks, where source
and target phrases are of length . is the num-
ber of blocks for for which
.
560
4 Online Training of Maximum-entropy
Model
The ... of Trigram language model: (c) probability
of predicting the first target word in the target clump
of
given the final two words of the target clump
of , (d) probability of...