QA for authoring and translation at Tekom tcworld 2012
I’m at Tekom tcworld 2012, in Wiesbaden. Yesterday I attended an excellent session entitled, “Addressing the Quality Gap: Unified Quality Assessment for Authoring, Human, and Machine Translation”. The presenters are Dr. Arle Lommel and Aljoscha Burchardt.
Here is the published summary of the session:
Language quality is difficult to asses and overarching standard methods for describing quality do not exist, especially across different tasks (human translation, machine translation, authoring). One of the goals of QT-Launchpad, a new project funded by the European Commission and led by DFKI, is to unify industry and research perspectives and requirements in a new, unified way of assessing quality that can be applied across many different tasks to allow comparison and benchmarking. Another project objective is the development of ways to automate the identification of translation problems and flag them for review and further attention.
I didn’t take notes during this presentation. Instead, I sat back and listened, and watched the rest of the audience too. People were very interested in what Arle had to say, as was I.
My main take-aways were:
- This is a very exciting and worthwhile project. It will greatly improve the current status of quality control in translation of technical documentation. The system now under development will provide standards and automation to make QA results transferrable and comparable across translation projects. The system will be flexible enough so that you can go with the defaults or configure the requirements and outputs to suit your needs.
- The project is currently in design phase. In January 2013 Arle and his team will invite input from stakeholders. If you’re interested in giving input, contact Arle Lommel.