Our Analysts' Insights

Based on press coverage of recent advances in machine translation (MT), it may seem as if all the major developments happen in the United States. However, since the 1980s, much of the basic research in this area has either happened in European projects or been headed by alumni of groups in the region such as the German Research Center for Artificial Intelligence (DFKI), University of Aachen, and the University of Edinburgh. Until 2016, the state of the art phrase-based statistical MT system was Moses, which was created in an effort funded by the European Commission, and research institutes in the region have been instrumental in contributing to new neural machine translation (NMT) technology that is rapidly advancing quality and capability. 

This long heritage of European investment in MT shows up in how readily LSPs there have adopted post-edited machine translation (MT). In 2016, our surveys of over 340 LSPs and more than 140 enterprise buyers of MT services found that 56% of PEMT providers in our sample were in Europe, compared to 26% in North America, and 18% in the rest of the world. Notably, four out of five of them have fewer than 100 employees, compared to about 70% elsewhere. 

This interest extends beyond LPSs: Enterprise implementers in Europe are more likely than their North American counterparts to adopt sophisticated deployment models that target high content volumes or large number of languages. Adoption tends to follow a predictable curve that requires high content volumes, but the points where MT makes sense are shifting dramatically to smaller and smaller amounts of text, a shift being driven by both large buyers of MT and LSPs in Europe. 
 

European Enterprises are More Likely than Those Elsewhere to Fall in the Turnaround Artist or Content Buster Categories

Because of the intense interest in MT in Europe, CSA Research teamed up with the E.C.-funded CRACKER Project to provide a close look at the landscape for machine translation in the EU and how both enterprises and LSPs in the region turn to it. The resulting study, which draws heavily on our 2016 survey of language companies that have adopted MT and our annual survey of the language services market, finds that Europe is the leader in this revolution, in terms of technology, business models, and adoption in the public sphere. The study finds both strengths and challenges to adoption in the region, and provides a focused look at Europe’s role in this area. 

As mainstream interest in artificial intelligence (AI) continues growing – and machine translation moves from a “maybe later” to a “must have” application – understanding the role MT plays for enterprises and the public sector can help guide investment and development decisions. This free report will help readers in Europe and elsewhere orient themselves in a market that is often dominated by hype and marketing claims. A clear view of this technology will help readers prepare themselves for the coming augmented translation revolution and find new business models that will help Europe – and other regions – meet exploding demand for content in the language of the information consumer.