MIMUG - Using Data to Increase Profits by Arle Lommel of CSA

The Multidimensional Quality Metrics (MQM) framework is a widely adopted hierarchical typology of translation error types that can be used to build custom translation quality evaluation metrics. Building on the legacy of the LISA QA Model, SAE J2450, and numerous other translation quality tools, it provides a comprehensive catalog of translation error types intended for use with both human and machine translation. With development funded by the European Commission (in the QTLaunchPad project), it serves as the basis for the TAUS Dynamic Quality Framework (DQF) error typology, and appears in numerous translation tools and enterprise evaluation practices. In this presentation, Dr. Arle Lommel, who led development of this framework, will discuss the rationale for it, why it is needed, and how both content creators and language service providers can apply it to meet their quality evaluation needs.

Dr. Arle Lommel has worked in the localization industry, with an emphasis on standards development, since 1997. Currently employed as a senior analyst at CSA Research, he worked for the Localization Industry Standards Association (LISA), the Globalization and Localization Association (GALA), and the German Research Center for Artificial Intelligence (DFKI). He led development of standards at LISA and spearheaded the creation of the Multidimensional Quality Metrics (MQM) framework for translation quality evaluation and chairs the ASTM working group that is continuing to refine it.