Challenges are there to be overcome. At the translation agency MondragonLingua we are firmly committed to advancing in the field of professional translations through responsibility and innovation. And always focusing on our clients.
Below follows a summary of the R&D projects in which MondragonLingua has participated or headed jointly with Vicomtech, its strategic technology partner. The results of the finished projects form part of the portfolio of services offered by MondragonLingua to its clients.
-PLATA (2012-14). In this project, in partnership with Vicomtech, MondragonLingua developed a machine translation platform for the translation of news and technical documentation. Also participating in the project were EiTB and IBM, to validate the statistical machine translation systems for news and technical documentation, respectively. This platform is the basis for the current MondragonLingua machine translation systems.
-IRTEN (2013-16). The aim of this project was the development of a comprehensive advanced management solution for international sales and communications processes. This project was headed by Informática 68 and involved companies such as Ibermática and Ibercom. The translation agency MondragonLingua contributed with the development of statistical machine translation systems and the development of an API for the integration of these systems with the IRTEN platform. This API is the one that is currently available for the integration of the different machine translation systems with other information systems (e.g. CMS, CRM, CAT tools).
-TRADIN (2015-17). The aim of this project was the development of customized machine translation systems for the industrial sector based on the intensive exploitation of data. In partnership with Fagor Arrasate and thanks to the development subcontracted out to Vicomtech, MondragonLingua has developed a total of four neural machine translation systems designed for industry. This project represented a qualitative leap in quality of machine translation systems, thanks to the use of neural networks.
These projects have been financed through the HAZITEK program of the Department for Economic Development and Infrastructures of the Basque Government.
-ADAPTA (2015-17). The purpose of this project was to develop technology for the development of fully customized machine translation systems based on the exploitation of heterogeneous data. In this project, MondragonLingua, in partnership with Vicomtech, has developed technologies for the generation of linguistic resources. These resources (generally previously translated texts) are essential for the development of machine translation systems. Thanks to the components developed in this project, MondragonLingua is able to reach a higher number of clients and offer its machine translation services to those who do not possess prior linguistic resources (e.g. translation memories). This project was financed by the RETOS program, organized by the Ministry of Economy and Competitiveness.
Also of particular mention is the TGATE project (2016-18). The aim of this project is the integration of machine translation systems with MondragonLingua‘s human translation processes and simultaneous integration with the complete flow of the translation process with the client’s external systems (e.g. CMS, CRM). It has the backing of LKS Infraestructuras for the deployment of the solution and Vicomtech for the back-end development. The results of the project will allow the integration of the neural machine translation systems developed by MondragonLingua into the value chain of the translation process. These results will provide MondragonLingua with the necessary flexibility to adapt to different client needs and configure different value propositions based on these needs. This project was financed through the HAZITEK program.
MondragonLingua sat on the industry consultative committee of the QTLeap project (2014-16), financed through the European Commission’s 7th Framework Program. The aim of this project was the investigation of technologies for machine translation based on “Deep Learning”. MondragonLingua contributed with its market vision in order to evaluate the potential of machine translation systems based on “Deep Learning” techniques as well as neural networks. Leading global organizations in the machine translation sector such as DFKI, the University of Prague, and IXA Group of the University of the Basque Country participated in this project.
The endorsement of a leading group
MondragonLingua forms part of the Engineering and Business Services Division of MONDRAGON.