Amazon Web Services Inc. today added support for 22 new languages to Amazon Translate, expanding the usefulness of a service that already powers a massive variety of applications for its cloud customers.

Amazon Translate is an automated translator that uses machine learning to convert text between languages. AWS provides it via a programming interface that allows companies to hook up the service to applications such as chatbots and customer support systems.

The 22 languages added with today’s update are Afrikaans, Albanian, Amharic, Azerbaijani, Bengali, Bosnian, Bulgarian, Croatian, Dari, Estonian, Canadian French, Georgian, Hausa, Latvian, Pashto, Serbian, Slovak, Slovenian, Somali, Swahili, Tagalog and Tamil. Amazon Translate’s AI algorithms are now versed in a total of 54 languages and dialects. In practice, that means the service can provide automated translations across some 2,804 language pairs, more than 20 times the number it supported a year ago.

The breakneck pace at which AWS is enhancing Amazon Translate is at least in part a reflection of the revenue opportunities the service presents. Automated translation is a specialized feature but one with cross-market appeal. Hotels.com LP uses Amazon Translate to localize hotel reviews submitted by users, while startup Protagonist Inc. is harnessing it for data analytics. In addition, Amazon certified the service for processing medical records last year.

Other use cases, like translating real-time responses generated by chatbots, have the added requirement of low latency. To address that, AWS will enable companies to run Amazon Translate closer to their applications by making the service available in six more data centers regions: U.S. West (Northern California), Europe (London), Europe (Paris), Europe (Stockholm), Asia Pacific (Hong Kong) and Asia Pacific (Sydney).

In parallel, AWS is working to enhance the theoretical foundations of its service. Researchers from parent Amazon.com Inc. have in recent years published multiple academic papers detailing new methods for performing AI translation. 

The company is investing even more resources to advance the broader field of natural-language processing, which is relevant for multiple AWS offerings besides Amazon Translation. The company has an entire blog for highlighting its researchers’ work in this area. In some projects, such as a recent effort to improve the quality of Alexa’s answers, Amazon collaborates with competitors such as Google LLC to tackle common areas of interest. 

Photo: SiliconANGLE