Free online translation services are improving thanks to better data and more sophisticated algorithms. Google's Franz Josef Och reports that Google Translate currently accommodates most of the world's translation, with more than 200 million users worldwide accessing the service each month.
Google Translate and Microsoft's Bing translator compute the most likely match between two parallel texts in a given language pair using statistical algorithms. However, Google Translate's output quality suffers in subject areas where translated text for parallel match-ups is not widely available. Although Google Translate cannot make a distinction between human translations and those produced by its own system, Och's team is working on programs for screening out bad machine translations. Also aiding Google's translation improvement efforts are users that translate text from their mother tongue, as they can use a new feature asking them to select the better translation among multiple options.
German Research Center for Artificial Intelligence's Aljoscha Burkhardt says a key advantage of linguistic rule-based machine translation is that it gives users more control over the outcomes. "You know what the rule-based system is doing, but you need experts for going into a new subject domain, which requires a new lexicon and a new way of constructing sentences," he says.
From Deutsche Welle
View Full Article
Abstracts Copyright © 2012 Information Inc., Bethesda, Maryland, USA