After nearly 30 years language technology has entered the machine learning era, according to Renato Beninatto with Nimdzi Insights.
Security was prepared but the line to get in to the session was long thanks to downtown San Jose traffic, bolstered by those coming to see the Overview of the Translation Technology Landscape from an Industry Perspective, presented by Renato Beninatto.
There is an abundance of technologies on the market today and together they account for less than $1 Billion of the $50 Billion-dollar language industry, the rest being human-based services. These humans at LSPs are working less and less with files, and more and more with APIs to systems that contain content. This is creating a problem in that only 10% of major web and software systems have a connector for language service providers to deal with.
The gap cannot be covered by increased file transactions as the need to have human translations continues to grow undaunted by the now titanic use of MT systems.
The proliferation of MT platforms, especially more recent Neural MT systems frequently include connectors to translation software. But this is not the case for most other major types of systems used today, as measured by the Nimdzi team. More than 90 percent of major systems remain disconnected from any automation in their translation supply-chain.
Emerging from the evolving technology landscape are some new problems and solutions, such as the needs for global sentiment analysis and monitoring, and new technology players aggregating features and services based on new industry needs.
Renato painted a clear opportunity space for new technology approaches, as this will not be solved simply by “better MT”.
My thanks to Janice Campbell and Adobe, IMUG, and Renato Beninatto, for putting on this heavily attended event. I’m looking forward to seeing them again at Loc World in Seattle soon and I hope to see you there as well.