E.g., 04/19/2018
E.g., 04/19/2018

Machine Translation Quality and Post-Editing

Machine Translation can help organizations save time and money, but the quality of the output must be considered. Learn about quality measurement and post-editing processes by checking out some of GALA's top resources.

Featured Content

MT Quality

Automatic Evaluation of Machine Translation: Moving Away from Word Matching Metrics
Lucia Specia (University of Sheffield / QT21)
The majority of MT evaluation metrics come from comparing MT to human generated content. Quality estimation metrics provide an alternative approach.

Comparing Errors of Neural MT with Errors of “Traditional” Phrase-based and Rule-based MT
Aljoscha Burchardt (DFKI / QT21)
Different MT types have various strengths and weaknesses. Learn what to expect from different methodologies.

Comparing Errors: Neural MT vs. Traditional Phrase-based and Rule-based MT
Aljoscha Burchardt (DFKI / QT21)
Different MT types have various strengths and weaknesses. Learn what to expect from different methodologies. Read the companion article to the popular GALA webinar (above).

A Language Approach to Machine Translation Quality Estimation
Juan Rowda (eBay)
What is MTQE? What is its purpose? How can you approach QE from a linguist’s viewpoint? Find answers to these and other questions.

Automatically Correct MT Errors from Human Post-edited Data
Marco Turchi (FBK / QT21)
Learn about Automatic post-editing (APE), a research area to fix MT errors by looking at corrections made by professional translators.

 

Post-Editing MT

Better, Faster, and More Efficient Post-Editing
Juan Rowda (eBay)
What’s the key to successful post-editing? Making only necessary changes, and avoiding preferential changes.

Understanding Post-Editing – A Practical Approach
Valeria Filippello & Andrea Stevens (SDL)
Learn what post-editing is, how to successfully integrate post-editing, and common post-editing scenarios for SMT output.

Effective Post-Editing in Human and Machine Translation Workflows
Stephen Doherty and Federico Gaspari (QT21)
Post-editing is an increasingly widespread activity, but there is a lack of training materials available. Learn different types of applications and the advantages of these methods.

Edit Distance and Post-Editing
Silvio Picinini (eBay)
Read about artifical neural networks, discover the history behind the neural hype, and consider expectations for the coming years in NMT.

Post-editing 2.0: Dynamic Quality Targets in the Post-editing Supply Chain
Lena Marg (Welocalize)
Explore the concept of “dynamic quality” and learn how this ties in with various levels of post-editing.

 

Additional Resources

Quality Translation 21 (QT21) 
Industry Initiative
Quality Translation 21 is a machine translation project which has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 645452.
Machine Translation
GALA Connect Group
This group is open for GALA members and provides an avenue for discussion of different MT approaches and developments.
randomness