268 search results - page 16 / 54 » Improving IBM Word Alignment Model 1 |
147
Voted
STOC
2000
ACM
15 years 8 months ago
2000
ACM
Multiple sequence alignment is a fundamental problem in computational biology. Because of its notorious difficulties, aligning sequences within a constant band (c-diagonal) is a ...
105
Voted
MT
1998
15 years 3 months ago
1998
We present a quantitative evaluation of one well-known word alignment algorithm, as well as an analysis of frequent errors in terms of this model's underlying assumptions. De...
136
Voted
ACL
2009
15 years 1 months ago
2009
Current tree-to-tree models suffer from parsing errors as they usually use only 1best parses for rule extraction and decoding. We instead propose a forest-based tree-to-tree model...
109
Voted
LREC
2010
15 years 5 months ago
2010
We present LIPS (Lexical Isolation Point Software), a tool for accurate lexical isolation point (IP) prediction in recordings of speech. The IP is the point in time in which a wor...
128
Voted
EMNLP
2010
15 years 1 months ago
2010
We propose a language-independent approach for improving statistical machine translation for morphologically rich languages using a hybrid morpheme-word representation where the b...