TAUS asked Lingotek's Robert Vandenberg, Lingotek's Vice President of Sales and Marketing about In-Q-Tel's announced investment in the Utah-based translation automation company. Why Lingotek and why now?
Government and defense agencies are interested in sharing data with each other, and expanding collaboration between agencies in the field of translation. They all have their own databases, but they are looking for synergy by pooling this data and moving towards advanced leveraging. We can provide them with tools to integrate and collaborate around translation. In-Q-Tel reviews 600 potential businesses a year before it makes an investment decision on only two percent of them, so this is a very powerful endorsement of our corporate project. Compared with our early investors, they will play a strategic role by opening up access to government agencies and acting as a contract vehicle for selling software. There has always been government funding for translation technology, along with improving transliteration tools and developing translation memories. But all these resources have been siloed and embedded, so what we can help these agencies bring them all together into a platform so that the intelligence people can analyze them.
Will Lingotek develop special tools for government translation services?
There are three aspects to this In-Q-Tel operation:
First, they are taking an equity share in our company, which allows us to develop and grow.
Second, they are licensing multiple enterprise versions of our software for use in government agencies.
And third, they will integrate a number of other language tools such as transliteration and MT systems for their specific needs.
The idea is to combine TMs with MT and recycle language data to train MT in a positive feedback loop, right across the agency map. This highly collaborative agenda is unique in my experience of government operations. It's not simply a case of helping an agency leverage their own data, but of sharing across multiple agencies, and possibly with allied agencies in the United Kingdom and Canada. I am amazed how far the US government is into the Web 2.0 mindset of massive collaboration. On the basis of this investment, we shall be going to market with a new tool towards the end of this year which will illustrate the possibilities of integrating content into the translation process in a community- and crowd-focused way.
With the recently announced Google Translation Center, the creation of the TAUS Data Association (TDA) and your own plans, the pieces are coming together for large-scale collaborative translation. Does this mean the competition for platforms is heating up?
Some aspects of the putative Google translation center are in line with and possibly inspired by Lingotek, but it is mostly about crowdsourcing - pushing out content for nonprofessionals to do the work. In this sense it is not really in competition with us, though you never know how things will develop. I think that the main focus will be on collecting good language data from human translations to improve the company's MT system. Our vision is not just about centralizing databases and offering web access, even though these are very important. We want to offer a unified suite for collaboration and translation that integrates TM, MT and any other productivity tool to help in the process. Our software platform is architected so that modules and web services can be added on progressively so that the whole system operates as a virtuous tool box. This will enable organizations to leverage not just their own data, but each others' data. The idea is to democratize translation. TDA is equally supportive of this concept. My only frustration is that we are not able to move faster towards more sharing.
