Machine Translation aims to produce high quality translations of content in a given language into a different language. Based on machine learning technology, modern machine translation learns to translate from data that expresses identical or similar content in different languages. Machine translation technology has become ubiquitous over the years, with millions of translation requests generated every day in a wide range of scenarios.
The technology can be adapted to specific domains and languages, after the appropriate selection and preparation of the necessary data. Optimal results are typically achieved with systems prepared to target specialised content, as they learn to produce accurate translations from domain-specific data. As a component, machine translation is an essential addition to create multilingual applications in a wide range of scenarios.
Core technology to train data-driven machine translation systems and translate text.
Automatic gathering of suitable corpora to train machine translation systems for specific domains.
Automatic document and sentence alignment from parallel and comparable data sources.
Data management and selection of optimal datasets to train machine translation systems.
Document translation that preserves structure and formatting.
Web-based API offering access to machine translation functionality.
Supported Languages: English, Spanish, Basque, Catalan (others in progress)
With constantly growing volumes of content in need of translation, language service providers aim to integrate accurate machine translation technology into their processes, in order to deliver high-quality translation services on time. Our solution produces highly adapted translation technology for user-defined domains, thus providing optimal answers to real-life translation scenarios and needs.
Subtitles are critical to provide accessible audiovisual content to all citizens. In a globalised world where language barriers need to be overcome, multilingual subtitling is a crucial component for an unrestricted access to global content. Our solution enables the generation of mutlilingual subtitles from human transcripts or from content generated via speech recognition technology, in offline or live mode.
Public administrations need to make their content available in the official languages of the country, as well as the languages of foreign visitors. The demand for multilinguality is thus high for the variety of topics and domains covered by public administrations. Machine translation helps answer this need for accurate translations of public information and expand access to public services across languages.
Every commercial sector is a world in itself, featuring characteristic vocabulary and expressions. Successful market expansion requires appropriate internationalisation processes that takes into account domain features and specificities. Our machine translation technology offers the means to support translation that is tailored to specific sectors. Specifically, it has been applied to the industrial sector where the effective management of multilingual information (technical operation manuals, maintenance procedures, etc.) is a requirement of the Industry 4.0 revolution in the globalized environment.
We have developed and transferred a machine translation solution to one of the largest Linguistic Service Providers in Spain. The solution enables the development and use of translation systems targeted to their clients’ needs and their translation processes.
We have developed machine translation systems for professional subtitling, in partnership with major actors in the sector, for offline subtitle translation and as part of a speech translation solution for live broadcasting.
We have developed and transferred a large range of machine translation systems adapted to particular domains, from highly specific industrial sectors to the broader news and entertainment domain.
|MT_CORE||Core functionality to train data-driven machine translation systems and translate text.||Machines & servers Linux|
|MT_CRP||Functionality to gather domain-specific corpora from public repositories.||Machines & servers Linux|
|MT_ALM||Functionality to align parallel or comparable documents and sentences.||Machines & servers Linux|
|MT_DSL||Functionality to select and prepare optimal datasets to train machine translation systems.||Machines & servers Linux|
|MT_DTR||Functionality for document translation that preserves structure and formatting.||Machines & servers Linux|
|MT_SRV||Functionality to access machine translation services via a web-based API.||Machines & servers Linux|