an online parallel Concordancer usable as a translation tool Cover Image

Paralleelkorpuspõhine tõlkeabisüsteem Internetis
an online parallel Concordancer usable as a translation tool

Author(s): Katrin Tsepelina, Kaarel Veskis
Subject(s): Language and Literature Studies
Published by: SA Kultuurileht
Keywords: Estonian; corpus linguistics; specialized lexicography; automatic lexicon generation; word alignment; translation software

Summary/Abstract: this article describes the creation process of an online english-estonian parallel concordancer that includes a bilingual english-estonian lexicon. the lexicon has been automatically extracted from a parallel corpus containing kDe software documentation and textual parts of kDe applications. Using the Uplug word alignment tool, we created six different versions of the lexicon that correspond to different preprocessing stages of the source corpus. the precision of the final versions of the lexicons was 96% for the english-estonian lexicon and 89% for the estonianenglish lexicon. We compared these different versions of the lexicon, using the common entries of all our lexicons and other english-estonian lexicons as gold standard. the result of our work is an online translation tool that will, as we hope, make a valuable resource for translators and software localizers.

  • Issue Year: LIII/2010
  • Issue No: 11
  • Page Range: 820-835
  • Page Count: 16
  • Language: Estonian