KantanMT API

(The KantanMT API is currently in BETA I testing – it will be released for wider testing in BETA II over the next two weeks)

The KantanMT API provides existing KantanMT users with a new way to interact with our service as an instant, on-demand translation offering, in addition to the job-based workflow that users have been familiar with to date.

The API leverages innovative technology developed by KantanMT to allow users to translate segments individually and in real-time, but still benefit from the high-quality, secure, and user-controlled translation service that KantanMT customers expect.  It uses the existing user client profiles listed on the KantanMT website to serve translations to client applications, so as users train and refine their client profiles over time, the quality of the translations returned by the API will reflect those refinements and allow users to retrieve much higher-quality translations than are possible with competing Machine Translation APIs.

The API uses the REST (Representational State Transfer) software architecture, meaning that customers can develop their own client applications or scripts and perform translation requests programmatically, using simple HTTP GET and POST requests as used by every web browser, and which are supported by most modern programming languages and frameworks such as cURL, Java, JavaScript, .NET, PHP and Perl. Multiple segments can also be bundled into a single HTTP request to reduce the number of requests required to translate user content. Customers can also develop API clients as plugins for their preferred CAT (Computer Assisted Translation) environment such as Alchemy Catalyst or SDL Trados if the application vendor provides an SDK (software development kit) for third party development.

The KantanMT API uses JSON as the messaging format for the responses it serves. JSON is a widely used and highly space-efficient format which enjoys built-in support in some programming languages like PHP and multiple third party parsing and encoding libraries for languages that lack native JSON support such as Java.

Perhaps the biggest benefit of using the API is the scope for completely automating content translation.  It is possible to prepare content for translation outside of KantanMT, submit hundreds or thousands of untranslated segments to the API, and perform post-processing on the translated content without any human interaction whatsoever.

However, the API is also well-suited to translating small to medium-sized volumes of content of the order of hundreds of segments. It is also useful when a user only wants to translate selected portions of a document or file rather than uploading all of it as is the case using the website interface. A final example of the API’s usefulness is the situation where a customer wants to review the translated output of individual segments immediately, as they are produced, rather than running a translation job on all the content and then reviewing the translations after the job completes.

Eric Chubb, Senior Software Engineer, KantanMT.com

For more information about KantanMT, or to sign up for trial membership please see www.kantanmt.com

4 Responses to “KantanMT API”

  1. “and perform post-processing on the translated content without any human interaction whatsoever.”

    Does it mean, the post-processing is done by a machine as well?

    • Hi Dmitry,

      Sorry if that was a little bit unclear. The post editing is done by human translators, however the entire process can be managed automatically.

      Eric

  2. Thanks for the write up!!

Trackbacks/Pingbacks

  1. The KantanMT API: Can You Translate Me Now? - June 13, 2013

    […] Eric Chub, a senior software engineer at KantanMT, says that the the API opens up the capability to translate in amounts that are as small as select portions of a document. However, as he writes in a blogpost, […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 33 other followers

%d bloggers like this: