CNGL research centre for global content technologies have been awarded more than €270,000 by Science Foundation Ireland (SFI). The funds will be used on three projects to progress research around personalized e-Learning, machine translation adoption, and the integration of computer-aided translation and machine translation.
iOmegaT – An Instrumented Replayable Computer-Aided Translation Tool – is one of the three projects and has been developed in close collaboration with global translation company, Welocalize, who contributed significant engineering and project management expertise.
The iOmegaT project provides a means of measuring the impact of machine translation on translator speed. This is done by recording the number of seconds per word required to translate from scratch or post-edit machine translated text at a sentence level. A big improvement on the current industry practice of asking translators to report on the time required daily to post-edit machine translation. iOmegaT produces hundreds of data points per day instead of just one and does not require accurate timekeeping from the translator.
The platform comprises an adapted open-source computer-aided-translation tool that is used as an editing environment by translators and a proprietary reporting component to record translator throughput and other data.
Based on an initial prototype developed by John Moran a CNGL Ph.D. student, iOmegaT was developed and initially deployed onsite in June 2011 in close collaboration with Welocalize. iOmegaT has been used by Welocalize to aid the initial selection of suitable machine translation systems for specific content and also to inform the negotiation of appropriate discounts for post-editing relative to purely human translation.
“iOmegaT brings intuitive, granular data to a recognised metrics gap and another dimension to the analysis of relationship between client content, machine translation quality, productivity and price” said David Clarke, Welocalize Principal Engineer for language tools.
About the funding
“The TIDA Award will enable us promote the acceptance and uptake of iOmegaT by translators and their managers”, says project leader Dr David Lewis of CNGL at Trinity College Dublin. “We will also develop complementary components that offer translation service managers who use the resulting data to analyse translators’ actions to both improve translation efficiency and to negotiate corresponding prices with clients”, adds Lewis.
The funding, announced by Minister for Research and Innovation Sean Sherlock, is part of SFI’s Technology Innovation Development Award program. This targets research projects with strong commercial prospects. The three CNGL projects are in the area of Digital Platforms, Content and Applications – key priority areas for the Irish Government.
“CNGL combines the best academic and industry minds to produce global intelligent content technologies with significant economic and societal impact”, says Professor Vincent Wade of Trinity College Dublin, Director of CNGL. “These awards again highlight the commercial relevance of our work and its potential in supporting Irish industry and creating high-value jobs.”
Already in 2013, CNGL spinout companies have raised €2.15M in venture capital, which will allow for the creation of 25 jobs for Ireland.