E.g., 04/03/2020
E.g., 04/03/2020

Centrum Lokalizacji C&M implements a new Statistical Machine Translation (SMT) system

Centrum Lokalizacji CM

Saturday, 24 May, 2014

Centrum Lokalizacji C&M is one of few companies in Poland offering machine translation (MT) services, delivered with the use of a proprietary system that is independent from global MT providers. Our unique solution comprises four co-existing components:

  1. A proprietary program for corpora storage and processing. It allows for corpus data extraction from TMX files, their division according to the disciplines, as well as their clean-up and export to a format supported by the SMT system training tools. The system stores corpora in a database, so they can be updated in a real time, linked freely, and easily managed.
  2. An open-source program for data preparation and translation model creation.
  3. An open-source program for SMT training and translation. The program is really fast and flexible; it enables translation in the phrase-based, hierarchical, and context-based mode. At its average working speed, it will process 30 sentences per second.
  4. A proprietary program for both pre- and postprocessing of the input and output SMT material. It is a rule-based mechanism that allows, among other things, to correctly process tags, capital letters, variables, and other untranslatable elements, as well as for various text modifications improving the quality of translation.

Working together, these four components produce machine translations of extremely high quality, especially in the field of IT and advanced technologies. We have also developed a system for quick evaluation of both “raw” and post-edited machine translations. All these solutions contribute to shorter turnaround time of services and reduction of costs without compromising the quality of translation.

For more information, please visit http://www.cmlocalization.eu/EN/

Centrum Lokalizacji C&M is one of the leading language service providers operating in the Central and Eastern European market. C&M is carrying out advanced works on machine translation (MT), human-aided machine terminology management and quality control tools. The company is a member of international and domestic trade associations, including GALA, TAUS Data Association, PSBT, and KOMTE.