TAUS launches Dynamic Quality Framework Tools

Machine translation is rapidly becoming a mainstream tool in the translation industry. Yet, there is little understanding of best practices for evaluating machine translation quality. There are no benchmark data on performance by language, industry, content type or technology used.

Inhaltsübersicht

Reading time: 01:03 minutes
TAUS has worked with its membership of experienced users to design the Dynamic Quality Framework Tools. This neutral and independent environment helps ensure members apply best practices for their MT evaluations, whether selecting a translation engine, measuring productivity or evaluating the final quality of translations. Members benefit from automated reporting and contribute to a platform for benchmarking of industry average performance. The DQF tools add to the Dynamic Quality Framework Knowledgebase and Content Profiling facility, providing a comprehensive range of resources to select and apply best-fit quality evaluation models for all translation scenarios. “We are very pleased to help members make more robust selections of machine translation technology, lower their evaluation costs and enable benchmarking so that there is greater awareness of performance in comparison to industry averages,” says Jaap van der Meer, director of TAUS. “Moreover, as a professional industry we need to measure and track the improvements we all make on the technology in order to validate the short and long-term investments.” “The industry can finally work together to tackle the translation quality evaluation challenges faced by everyone. The TAUS Labs team will now focus on developing the DQF Tools to help improve the correlation between automated and human evaluation metrics, further magnifying the benefit of the TAUS Dynamic Quality Framework,” says Rahzeb Choudhury, TAUS operations director. <link http: www.tauslabs.com dynamic-quality about-dqf>www.tauslabs.com/dynamic-quality/about-dqf