Translate forums: Language Technology will incorporate

Representatives from Europe's leading language technology companies met in Brussels on 8th-9th October and agreed to form a legal entity representing their interests. The main outcome of the Workshop, attended by over 50 representatives, was the decision to set up LT-Innovate as a legal entity to serve as representative body for LT vendors at European level. This decision comes in the wake of the very successful LT-Innovate Summit that brought together more than 160 LT stakeholders in June 2012. Strengthening and raising the profile, as well as serving as a lobby to speak as one voice with industry and government is seen as increasingly necessary by many language technology companies, many of them SMEs. In a world and EU brought increasingly together by economic forces and Internet, the language technology industry must be an enabler and not a bottleneck in communication, thus opening new markets and business areas. Even within the EU itself, the situation is far from ideal. For example, Nikiforos Diamandouros, the European ombudsman, has criticized the European Commission for refusing to conduct public consultations in all the 23 official languages of the European Union. In a decision timely published two days after the event (11th October), Mr. Diamandouros found that the Commission's practice of launching public consultations in only a few EU languages constituted maladministration. Day 1 concentrated mainly in the provision of information about the new funding opportunities by the EU on Language Technologies and the direction future efforts will take. Several technology developers, like Pangeanic, proposed their ideas in order to obtain feedback from peers and EU representatives.

Day 2 concentrated on mission statement and organizing committees and focus by the key players on the European Language Industry, from machine translation to speech recognition and other areas, focusing on an SME perspective. Joachin Hummel stated that “Only by organizing the industry, and presenting the voice of 100 companies or so to decision makers, we can put enough pressure to make our voices heard to those who have to take a decision on funding.” Rubén Riestra intervened later, presenting on “Developing a Vision Statement for the Industry”. Sadly, he said, it is easier to move an apple through our borders than content through our digital borders. The EU does not have the military muscle to pull from the economy like the US has done nor the massive labor force resources like China. "We are flexible and we have a vision." His vision is that “the European LT industry will build a language-neutral Digital Single Market before the end of the decade. If we do not address this, other players will come and take the market." "Funding agencies will fund results and not efforts. We need to build an infrastructure that others will take advantage of (like the roads that transport apples and cable companies and other industries use)." He continued by saying that "Europe needs to build an industry, in this case, the Language Industry, establishing legal frameworks and policies" by

lowering barriers
creating economies of scale
fostering skills

Technology is an enabler, not a solution in itself. Many LT companies find it difficult to find marketers and sales personnel to sell their technology easily as a plug-in, an enabler to other industries. LT needs to work together with other industries. Rose Lockwood intervened later, presenting on "From niche to pervasive." Rose began by saying that new ICT ecosystem is affecting Europe’s ability to compete. Before, there was a hardware and software market, a telecoms market, etc, but the combination of platforms and data applications and the barriers are increasingly fuzzy.

Now we have a first layer composed by hardware companies like Cisco Samsung, Alcatel, Ericson, Nokia, a second layer with network operatos like BT, Deutsche Telecom, Telefónica, Vdafone, but it is the layer 3 we need to observe clearly as the include the big data players and these range from BBC to Google, Microsofot, etc and all massive digital publishers, of course now including video (Youtube, etc). That’s where EU fails. However, ICT firms in EU are not less R&D intensive (typically more) but they are

not specializing in new ICT sector, innovating in the less dynamic parts of the ecosystem and missing out completely on new segment of the market
concentrating investment in Layer 2 (least R&D intensive in the ecosystem)
not generating new firms that become leading innovators

Thus, Europe has a problem with capturing value in Layer 3 markets (Breugel) because

a) Digital market is not integrated b) Language barriers c) Absence of ICT clusters that reduces synergies.

HOWEVER: Europe’s cultural differences could also be an opportunity to differentiate and create niches, conditional on being able to each critical scale. LT is a new ICT market that can unlock potential for Europe because it is founded on decades of European R&D with world-class levels of technological expertise. It also addresses linguistic barriers directly, with multilingual and translation capabilities. Furthermore, it enables advanced communication with intelligent natural interfaces. It is the only technology capable of handling the massive amounts of data we are facing today. The size of our industry is approximately 20B € worldwide growing at 11% a year. There are 3 segments in the industry: intelligent content, translation and speech technologies. The first is overwhelmingly US-based, Europe takes the lead in the second, whereas speech technology is quite level even with the rest of the world. The world is moving from a fee-based, charging per licenses, to a model where enterprises monetize through usage fees. So, where can innovation happen in Layer 3? In Data, Application and Platforms and across many industries, for example, LifeScience /Healthcare will need Clinical Data with an application in the Public Health Monitoring, needs a platform to do so. The innovation is going to come from platforms oriented to a significantly large enough domain and the industry has enough room to succeed, as long as it is multidisciplinary. For example, in the shape of a High Quality Machine Translation (HQMT) platform, a self—service voice cloning platform, etc. It is going to come from applications based on integrated LT features which are likely to be vertical industry or domain specific, although there are also opportunities for existing applications replicated for new markets (speech components for new languages, analytics services, etc). Phillipe Wacker spoke later, calling for the generation of an industry vision. Today, niche technologies are in silos and our next step is to have a single industry with a sense of belonging together and a common vision/map for the way forward. Firstly, LT must be the cornerstone of the Digital Single Market: the European Language Cloud (infrastructure). Secondly, LT as a pervasive key enabling technology for vertical market segments in which Europe has global competitive advantage which can be consolidated and reinforced. In order to build the ecosystem, demand has to come from the buyers, who state their needs. It seems that nowadays, vendors have solutions that cannot market or for which there is a no market!! Researchers are the engines that facilitate the whole process. The mixture of the 3 creates innovation engines where we can identify and catalyze vertical innovation value-chains. LT-Innovates will provide a forum for buyers and vendors of LT technologies to discuss their needs and thus guide future R&D. The agenda is to articulate/promote vis-à-vis other stakeholders (researchers, buyers, investors) and policy makers. Within the next few months, the organization will pursue several objectives identifying innovation opportunities for its members, initiating collaborative projects addressing opportunities, removing inhibitors and implementing solutions. Finally, it also envisages the organization of a Buyer Focus Group. The decision was made to incorporate LT-Innovates as a legal entity, creating a Company Limited by Guarantee, based in the UK, which will have not-for-profit status. Phillipe's full presentation has been made available here.

Next time you think languages, think Pangeanic Machine Translation Engines from PangeaMT

Carla Estefanía

10/19/12