If Urdu/English machine translation combination is accurate up to 85%, what is stopping the development into other, more common pairs? In our...
Pangeanic collaborates on cutting-edge EU machine translation post-editing project
As part of Pangeanic's machine translation division, PangeaMT is collaborating with the EU program CASMACAT, a next-generation workbench being developed by leading machine translation and pattern-recognition authorities. Key contributors include Prof. Philipp Koehn from the University of Edinburgh, Prof. Francisco Casacuberta from Valencia Polytechnic, and Prof. Michael Carl from Copenhagen Business School.
"I'm very excited to align Pangeanic's experience with the cream of EU research for the next generation of machine translation tools and workbenches. CASMACAT is an example of machine translation as a productivity enhancer," said Alex Helle, Pangeanic's Head of Machine Translation Research. "Furthermore, initial research shows that it is possible to integrate the best of statistical reporting and confidence scores, plus translators' aids into a single, useful workbench," Mr. Helle continued.
Enhancing translation capabilities with language testing and real-world scenarios
Pangeanic will contribute its expertise by testing languages with little or no affinity to English, such as Japanese and Chinese, in addition to Spanish. Particular attention will also be paid to morphologically-rich languages like German. The project will also ensure compatibility with real-life translation scenarios, where companies need to translate format-rich content, including various tag types and XML flavors.
"We have already started experimenting with the first sets of data in Japanese. We look forward to checking the tests and working on further experimentation with the project’s algorithms so that the top of European research will enhance our own machine translation service product lines at PangeaMT and provide more efficient and user-friendly post-editing methods," added Mr. Helle.
Bridging the gap between technology and human translators
While the benefits of machine translation to accelerate publication and translation processes are widely accepted, post-editing and technology adoption by translators remain significant challenges for many companies. CASMACAT aims to address this by offering interactive methods that engage translators through several post-editing techniques, such as confidence score color identification and suggestion features.
Pangeanic’s role involves testing the latest research in actual translation workflows, providing feedback to researchers, and customizing a version of the machine translation workbench.
Learn more at CASMACAT’s User Group homepage: http://www.casmacat.eu/index.php?n=Main.UserGroup
Related Articles
The MosesCore consortium, an EU Project aimed at promoting open source machine translation, is sponsoring the Machine Translation Marathon 2012.
Pangeanic leads a consortium with leading machine translation and NLP companies KantanMT and Tilde, to create the largest-ever neural machine...