Maximum Likelihood Alignment of Translation Equivalents

We describe a corpus-informed lexical acquisition procedure based on maximum likelihood estimate of translations. The most likely translation words in singleton parallel sentences are determined by the measure of the similarity of their distribution in the entire corpus. The results show that for th...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
1. Verfasser:	Amsalu, Saba
Format:	Tagungsbericht
Sprache:	eng
Schlagworte:	Applied sciences Artificial intelligence Computer science control theory systems Exact sciences and technology Machine Translation Optimal Alignment Parallel Corpus Source Language Speech and sound recognition and synthesis. Linguistics Target Sentence
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	We describe a corpus-informed lexical acquisition procedure based on maximum likelihood estimate of translations. The most likely translation words in singleton parallel sentences are determined by the measure of the similarity of their distribution in the entire corpus. The results show that for the recall level obtained our procedure is quite efficient.
ISSN:	0302-9743 1611-3349
DOI:	10.1007/11816508_47