Language detection, language translation... it IS exciting, so why I am yawning?

by Michael S. Kaplan, published on 2008/03/21 17:36 +00:00, original URI:

Content of Michael Kaplan's personal blog not approved by Microsoft (see disclaimer)! In particular, I am speaking for neither Microsoft nor Google nor AltaVista nor Yahoo nor Systran nor my nieces nor my grandmother nor any ex-girlfriend I have ever had nor anyone else, other than me. Those are the rules -- learn 'em, know 'em., love 'em....

Over on The Official Google Blog, Product Manager Brandon Berger just blogged New Google AJAX Language API - Tools for translation and language detection.

Now this kind of thing is exciting, though as usual the terms of use over there keep me from getting into it too much.

The documentation and samples are here, which contain answers to the all-important coverage. Here are the supported languages (Google Translate is reported to have the most up to date list):

And the supported language pairings are:

which will make German and French users slightly happier than most, but which I guess proves that the desire to make English the most important language is important at Google just like at Microsoft. :-)

I guess one of the only thing that disturbs me here is that these lists are not very specific about how much is related to detection rather than translation. Given the piss-poor support by any company of actual auto-translation, I find the notion of adding language detection to be incredibly appealing and would have liked a lot more in the way of detection samples and maybe even a little on the roadmap for that feature in the future....

From the point of friendly advice to a rival company, the language pair list is the same one that has been on Google Translate all along -- so although this is a pretty exciting feature announcement (Google Translate has been around for ages but reverse engineering that page to call it yourself does violate the terms of service so a supported API is indeed a compelling feature to stay out of legal hassles!), I am left feeling a little cheated -- like they just decided to spin off and let developers do what they were exposing anyway without adding new languages at the same time. In fact if you ignore the Arabic support, the Systran list (used by both Alta Vista and Yahoo) which has been around much longer is equally compelling but also pretty stagnant in terms of improvements.

Shouldn't Google be trying to wow me here a bit more by having a lot more than the competition rather than by trying to show they made the lawyers back down?

Now Microsoft does not have product out there of its own that can claim any better, so this is not one of those "I think Microsoft is better" things. I just worry that the message people will get from this stuff is that the other languages are too hard to do so no one wants to try?


This blog brought to you by 󠀁 (U+e0001, aka LANGUAGE TAG)

comments not archived

Please consider a donation to keep this archive running, maintained and free of advertising.
Donate €20 or more to receive an offline copy of the whole archive including all images.

go to newer or older post, or back to index or month or day