This complimentary report comes at a critical moment as two prerequisites for accelerated innovation and greater interoperability, industry-wide language data sharing and open translation platforms, have become reality. This report outlines the industry's development focus in the coming years, highlighting the types of decisions companies are making, spotlighting a few case histories and predicting how events will unfold.
Underpinning this research is the assumption that there is a fundamental and positive shift from business and innovation models based on control and hierarchy to ones empowered by collaboration and openness.
The report represents the industry's collective wisdom based on survey data, group discussions during the TAUS World Tour, 30 in-depth interviews, and TAUS articles and reports.
If you are interested presenting at the TAUS User Conference, Portland, Oregon, USA, 27-30 October 2009, this report should be used as a guide to help shape your proposal.
From crisis to opportunity
As you would expect the market survey confirmed the global economic slowdown has led to less translation volume (49% of respondents) and less languages being translated (20%). However, whereas some other sectors have been devastated, the picture for the translation industry is not dire, with 20% reporting no slowdown and 36% exploring new innovation opportunities.
MT becomes mainstream
The headline finding is that over 80% of respondents expect to be using Machine Translation (MT) as part of the mainstream of business within two years, up from 37% today. Counter intuitively, only half of respondents plan to share language data, a pre-requisite for improving MT output quality, which is considered the main barrier.
We expect more MT providers to move to service oriented business models, and form partnerships with academic institutions and language service providers. We note that there are clients and providers building on top of open source MT engines, fuelling this movement.
Language data sharing revolution
As technical constraints are overcome, early adopters of language data sharing will rapidly gain cost advantages. When sharing takes off, barriers to entry for MT will come down, more MT languages will become available, and domain specific engines will raise quality. This will improve the competitive environment and super charge the innovation landscape. It will become much easier for entrepreneurs to tap into unmet needs for instant, portable, and personalized translation.
The crowd business
The hype around community translation is settling down and the companies we spoke with have been creative, industrious and level-headed with the crowds that they have been working with. Eighteen percent plan to be using community translation within two years, a reflection of the limits to what volunteers can offer to the business of translating commercial material. Using community translation to improve translation memories and train MT systems is undoubtedly the largest long-term benefit to the industry.
Reinvention
The take up of open translation platforms is a positive development, whether openness is defined in terms of enabling seamless connectivity between systems, harmonized standards so that there is portability of information and/or open source for fostering more collaborative innovation. Recent announcements by large, small, old and new companies on plans for more open systems highlights a general trend away from lock-in business models. A quarter of suppliers could be using such systems within two years. This points to a general shift towards software-as-a-service based models for language service providers and is the clearest sign yet of progress for an industry reinventing itself.
Open for continuous
Over a third of clients are aiming to be working in a continuous translation environment within two years. This will reinforce advantages for providers with open translation platforms, which offer clients much needed interoperability, and drive down customization costs.
Converging with customer support
Companies are eager to find cost effective models to localize support content. Human translation, whether by professionals or a community, is only workable in a limited number of cases. Improved MT offers the potential to tap into a huge opportunity to provide localization for Knowledge-Centered Support.
A new agenda
Translation personalization remains largely a research objective among academic circles. Two ingredients for making this happen on a bigger stage, industry-wide language data sharing and open translation platforms, are now reality. Success would mean hugely disruptive innovation and result in world changing impact.
ARTICLES ON OTHER REPORTS IN 2009
LSPs in the MT Loop: Current Practices, Future Requirements (July)
ProMT: Machine Translation Solutions Tailored to All Types of User Environment (July)
Owning MT. Lionbridge and SDL as MT users (April)
Microsoft's Machine Translation (March)
Technical Guide to Statistical Machine Training Data (March)
Symantec's Localization Revolution (January)