Difference between revisions of "Earley-based structural transfer for Apertium"

From Apertium
Jump to navigation Jump to search
Line 10: Line 10:
 
==Further reading==
 
==Further reading==
   
* [http://66.102.9.104/search?q=cache:GwOQsQtddJIJ:www.isi.edu/natural-language/mt/hlt-naacl-06-zhang.pdf+earley%27s+algorithm+machine+translation&hl=en&ct=clnk&cd=3&client=iceweasel-a http://www.isi.edu/natural-language/mt/hlt-naacl-06-zhang.pdf]
+
* Koichi Takeda [http://66.102.9.104/search?q=cache:E-giecHfU8QJ:acl.ldc.upenn.edu/P/P96/P96-1020.pdf+earley+algorithm+%22machine+translation%22&hl=en&ct=clnk&cd=6&client=iceweasel-a Pattern-Based Context-Free Grammars for Machine Translation]
  +
:This paper proposes the use of "pattern-based" context-free grammars as a basis for building machine translation (MT) systems.
* [http://66.102.9.104/search?q=cache:PZewJi8kmc0J:www.slt.atr.jp/IWSLT2006/proceedings/EC_13_NTT.pdf+earley+algorithm+%22machine+translation%22&hl=en&ct=clnk&cd=5&client=iceweasel-a http://www.slt.atr.jp/IWSLT2006/proceedings/EC_13_NTT.pdf]
 
*[http://66.102.9.104/search?q=cache:E-giecHfU8QJ:acl.ldc.upenn.edu/P/P96/P96-1020.pdf+earley+algorithm+%22machine+translation%22&hl=en&ct=clnk&cd=6&client=iceweasel-a http://acl.ldc.upenn.edu/P/P96/P96-1020.pdf]
+
*Randall Sharp and Oliver Streiter [http://66.102.9.104/search?q=cache:QJHB-s5Ze8cJ:www.iai.uni-sb.de/docs/meta93.pdf+earley+algorithm+%22machine+translation%22&hl=en&ct=clnk&cd=7&client=iceweasel-a Simplifying the Complexity of Machine Translation]
*[http://66.102.9.104/search?q=cache:QJHB-s5Ze8cJ:www.iai.uni-sb.de/docs/meta93.pdf+earley+algorithm+%22machine+translation%22&hl=en&ct=clnk&cd=7&client=iceweasel-a http://www.iai.uni-sb.de/docs/meta93.pdf]
 

Revision as of 13:45, 20 May 2007

Perhaps [Earley's algorithm http://en.wikipedia.org/wiki/Earley's_algorithm] to parse context-free grammars (which has a left-to-right longest-match philosophy as Apertium) could be used to perform more complex syntactical transformations; this could be useful for distant language pairs containing embedded structures.

Open questions:

  • Currently, Apertium uses text streams to communicate. I assume this would not be possible here.
  • When would one call the bilingual dictionary? Apertium Level 2 calls it in the first stage.
  • We should check whether this has been done before.


Further reading

This paper proposes the use of "pattern-based" context-free grammars as a basis for building machine translation (MT) systems.