SYSTRAN'S translation technology

What is machine translation?

Machine translation, or automated language translation, is a process using computer software to automatically render a text presented in a given natural language (such as English) into another language (such as Spanish). Applying complex sets of specific rules, the software analyses and then transfers the grammatical structure of the source language (text to be translated) into the target language (translated text).

Machine translation is not a simple process. It does not involve the mere translation of each word in any selection of text but rather the rendering of all the words in the given sentence or context in the target language, achieved through the understanding of the entire passage. Natural languages are distinguished by their morphology (fundamental elements and word formation rules), their syntax (sentence structure) and their semantics (meanings of words). They also involve a great number of ambiguities.

SYSTRAN, a machine translation pioneer

SYSTRAN was one of the earliest developers of machine translation software. In 2002, SYSTRAN released a completely redesigned architecture, leveraging more than 30 years of enhancements that make its translation technology more efficient.

Combined with the increased performance offered by today's computers, SYSTRAN now provides greater translation coverage.

A vast linguistic database to ensure the highest possible translation quality

SYSTRAN's translation software products and solutions are developed on Linux and run on Microsoft Windows and all Unix platforms. They are all built on the same translation engine and make use of the latest natural language processing technologies.

SYSTRAN's translation systems integrate finite state technology to ensure fast access to very large linguistic knowledge bases. SYSTRAN makes use of a broad range of linguistic resources so as to guarantee precise and rigorous translations.

Continuing improvements in machine translation technology

Standards such as XML, Unicode and HTTP are fundamental to the multilingual objectives served by SYSTRAN's technology: Web-based applications, e-mail and instant messaging, intranets, content management, etc.

SYSTRAN's R&D team keeps abreast of the latest advances in linguistics and natural language processing, focusing resources in areas with the greatest potential. Its aim is to continually develop new machine translation systems to improve translation quality and the robustness of solutions, built using a multi-agent, multi-threaded architecture.

White papers