Quality Estimation Insights
André discusses the challenges and improvements in quality estimation, emphasizing the importance of tagging words in machine-translated sentences as good or bad based on post-edited translations. The goal is to evaluate machine translations without using a reference, focusing on fluency and adequacy to enhance translation quality.In this clip
From this podcast

NLP Highlights
58 - Learning What’s Easy: Fully Differentiable Neural Easy-First Taggers, with André Martins
Related Questions