Knowtions aims to make translating complex scientific documents into any language faster and more accurate
Companies wishing to translate regular business documents into other languages can easily find a number of professional services and software tools online. But for those that need to translate highly complex materials on a specialized topic – a medical publisher, say, or pharmaceutical company – the task is much trickier. Not only do you need a translator who’s fluent in multiple languages, you also require someone with an expert understanding of the science you are translating.
That’s the opportunity Christina Cai (BA 2014 New) and Anthony Lee (BA 2015 Woodsworth) see for their startup, Knowtions. Since founding the company in 2013, Cai and Lee have recruited some 500 PhD-trained subject experts, many of whom are fluent in multiple languages, as part-time translators. About a third of these experts are based at U of T.
Knowtions’ custom-built software allows multiple experts to work collaboratively and in real time regardless of their location, which, Cai says, speeds up service to clients and ensures greater accuracy in the finished translation. At the same time, the Knowtions system stores information regarding the usage and context of the highly specialized terms in the documents they are translating, which makes future translations including these terms faster.
The system is notably people-driven, though: “Most software is extraordinarily bad at translating words it doesn’t recognize. The problem is especially acute with languages such as Chinese and Japanese, where there are no spaces to delimit what constitutes a word,” says Lee. Among the many inefficiencies Knowtions is striving to eliminate is “one of the most time-consuming aspects of a translator’s job: terminology research,” says Cai. As the company’s database of complex terms and jargon for a variety of languages grows, translation becomes faster and simpler.
Knowtions’ team of experts would have no problem translating a sentence such as “Explain the role of TNF-alpha and IL-1 in bladder cancer,” but it would stymie most common machine translators, which do not have the data to translate words specific to fields such as neuropsychiatry, optogenetics and other scientific disciplines.
Knowtions, which currently works with small to medium-sized businesses, is based at U of T’s Banting and Best Centre for Innovation and Entrepreneurship, where Cai and Lee receive strategic advice from leaders in the field. Now, with seed funding, the company plans to expand beyond life sciences, into law, engineering and aeronautics, and to seek work with large medical and scientific publishers and pharmaceutical and medical device companies. Further into the future, Lee hopes to license Knowtions’ translation engine and data to companies such as Google, Baidu and Microsoft.
Ultimately, Lee sees the company as a source for answering difficult scientific questions in any language, using its trademark combination of human and machine intelligence. “That’s our vision,” he says.