An integrated solution: applying PROMT machine translation technology, terminology mining, and the TRADOS TWB translation memory to SAP content translation

An Integrated Solution: Applying PROMT Machine Translation Technology, Terminology Mining, and the TRADOS TWB Translation Memory to SAP Content Translation Dr. Ulrich Boehme, Svetlana Svetova Paper type: U Dr. Ulrich Boehme SAP AG Neurottstrasse 16, 69190 Walldorf Germany ulrich.boehme@sap.com Svetlana Svetova PROMT Vyborgskaya emb., 61, St. Petersburg, Russia svetova@promt.spb.su Abstract This paper describes the experiences of SAP and PROMT specialists with applying the PROMT English-Russian machine translation system, the PROMT Terminology Manager Tool for automatic terminology extraction, and the TRADOS TWB translation memory system to the automated process of translation of SAP content from English into Russian. Keywords English-Russian machine translation system, automatic terminology extraction, automated translation process. Introduction SAP AG is one of a few German companies that are investing in Russian translation and localization. SAP started translating its products into Russian 10 years ago and is now able to provide the Russian-speaking countries in the CIS with 15 software products as well as documentation and training courses. As a result of translation, SAP also provides its customers with extensive Russian terminology: about 35,000 entries in the SAP terminology database STERM and more than 50,000 entries in SAP’s electronic dictionary, provided by PROMT. The volume of documentation and training courses translated into Russian so far is about 8 million words. If SAP’s Russian market continues to grow steadily, we can expect to translate an even larger volume of knowledge content. Knowledge content includes SAP’s implementation guide, SAP release notes, selected SAP industry solutions, and mySAP.com documentation. These can be accessed via the SAP Help Portal. It is very difficult at the moment to forecast the exact volume of words or lines not yet translated into Russian. It is also clear that SAP Language Services’ (SLS) budget for translation of the above-mentioned text types will increase dramatically. Therefore, SLS has been forced to look for efficient technology-supported translation into Russian to reduce time and costs of future translation projects. >From the very beginning, SLS was looking for an integrated solution that combined translation memory technology (TM), machine translation (MT) technology, automatic terminology extraction technology, and, if possible, (even simple) text control mechanisms for the source texts to be translated. The first step towards reducing time and costs for translation was implementing translation memory technology, namely TRADOS TWB, in 1996. However, the unexpected increase in entirely new texts (No matches) has forced SLS to look for a commercial • MT system (English to Russian) that Runs on Pentium PC with MS NT/Windows and MS Office • Has development potential and is maintained by the software developer • Is easy to use After successfully implementing TRADOS translation memory technology, SLS evaluated several English to Russian machine translation systems. • The requirements were as follows: Provide steady results from machine translation while processing large volumes of documents with acceptable quality • Allow easy integration with TM systems • Provide easy-to-use, effective tools to extract, process, and manage terminology, and also to store extracted terminology in the user dictionaries for machine translation • Reduce …

Content truncated to 500 words