Jurnal Teknik Informatika Volume 04 Nomor 01 Maret 2011 ISSN : 1907 – 9834
Kharisman Kholid Hudaya
Jurusan Teknik Informatika – STT Nurul Jadid Paiton
Example Based Machine Translation (EBMT) is a translation paradigm that use example translation corpus source. This paradigm is the newest than two paradigm exist before: Rule Based Machine Translation and Statistical Machine Translation. The accuracy of output translation from EBMT is depend on completeness coverage example corpus translation. Even with a reasonably large corpus, we will not have examples that cover everything we want to translate.
The quality of ouput sentence produced by machine translation often out of expectation. Therefore, Computer Aided Translation more eligible because human has more control to classify a word to a contextual meaning. This research had a purpose to develop a Computer Aided Translation Arab-Indonesia prototype with EBMT paradigm adaptation, extended by Tashrif System as a proposed new method. Experiment on domain Fiqh Islam get a total average precision 73% and 42% on Arabic news website Al-Jazeera.
Keywords: EBMT, Computer Aided Translation, Arab-Indonesia, Machine Translation