| | |
| | |
Stat |
Members: 3645 Articles: 2'501'711 Articles rated: 2609
20 April 2024 |
|
| | | |
|
Article overview
| |
|
Statistical Machine Translation by Parsing | I. Dan Melamed
; Wei Wang
; | Date: |
2 Jul 2004 | Subject: | Computation and Language ACM-class: I.2.7 | cs.CL | Abstract: | Designers of statistical machine translation (SMT) systems have begun trying to exploit tree-structured syntactic information. This article offers a coherent algorithmic framework to facilitate such efforts. Our main contribution is a generalization of the common notion of parsing. In an ordinary parser, the input is a single string, and the grammar ranges over strings. In order to use syntactic information, an SMT system requires generalizations of ordinary parsing algorithms that allow the input to consist of string tuples and/or the grammar to range over string tuples. Three particular generalizations, connected by some trivial glue, are all that is necessary for syntax-aware SMT: A synchronous parser is an algorithm that can infer the syntactic structure of each component text in a multitext and simultaneously infer the orrespondence relation between these structures. When a parser’s input can have fewer dimensions than the parser’s grammar, it is a translator. When a parser’s grammar can have fewer dimensions than the parser’s input, it is a synchronizer. This article offers a guided tour of these generalized parsing algorithms. It culminates with a recipe for using generalized parsing algorithms to train and apply a syntax-aware SMT system. | Source: | arXiv, cs.CL/0407005 | Services: | Forum | Review | PDF | Favorites |
|
|
No review found.
Did you like this article?
Note: answers to reviews or questions about the article must be posted in the forum section.
Authors are not allowed to review their own article. They can use the forum section.
browser Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; +claudebot@anthropic.com)
|
| |
|
|
|
| News, job offers and information for researchers and scientists:
| |