This is the guide about Analyze Files report in SDL Studio Trados 2019, that help you to use SDL Studio Trados 2019 easier.
This topic explains what some of the data in the Analysis report means, more specific, data regarding upLIFT and AdaptiveMT.
- Fragment words (whole TU) – the number of whole TU fragments found. A whole TU fragment is a complete TU found within a longer segment that would normally be considered a fuzzy match or lowe in the Analyze Files report.
- Fragment words (TU fragment) – the number of fragment words found. This is the number of times relevant fragments are found in a segment. This can provide an indication of how useful a TM can be even if there are no fuzzy or 100% matches
- AdaptiveMT Impact – this percentage is an estimation of the reduction in post-editing for each of the segment type row. To create this percentage, the baseline MT engine is taken into consideration, then an assessment is made based on the changes brought by the AdaptiveMT engine. This percentage thus shows how much use you are getting from your AdaptiveMT engine. If you have several AdaptiveMT engines in use, you will be able to see which one will give the best result for a particular project.
- New/AT – when you do not use automatic translation, this word count displays all new words. When you use automatic translation, this word count displays automatically translated words.
- AdaptiveMT Baseline – this word count displays how many words/segments have been handled with machine translation based on the baseline engine.
- AdaptiveMT with Learnings – this word count displays how many words/segments have been handled with machine translation based on the adaptive engine. This as an indication of how much value you get from your adaptive engine.