What is aligning?
Aligning is the process of pairing the source file segments (untranslated) with target file segments (the source file translations). During the process, the output XLSX file is created with segments pairs. The output file, after small modifications (deleting proper columns and rows in the file), can be used as translation memory and be imported to XTM.
Please, keep in mind that TM Align is not available on freelance XTM and group accounts. You need align access rights to use these features! |
|---|
Aligning process
In order to see a step-by-step aligning guideline, please, familiarize yourself with the respective article: How to align translated documents.
The process is started by uploading the source files and choosing the options:
After the process is started, the analysis of the input files is performed in order to create XLF files on the back-end side that are used for alignment.
Once the alignment has been performed, the output files are created which are downloadable in XTM UI:
Alignment related files
The downloaded zip file always contains two Excel files:
“Aligned.xlsx” contains all aligned segments;
“Aligned.90+.xlsx” contains only the segments with the match probability value higher than 90%.
It is recommended to review only the aligned.90+.xlsx document to quickly create a high accuracy translation memory.
aligned.xlsx
This file contains all the segments from the input files with some additional data:
The file's content:
Src Seg. (coulmn “A”) → the number of the segment from the source file;
en_NT (column “B”) → source segments language code;
fr_CH (column “C”) → target segment language code;
Tgt Seg. (column “D”) → the number of the segment from the target file;
Probability (coulmn “E”) → probability for the target segment to be the translation of the source segment;
prevCRC (column “F”) → CRC value of the previous segment;
nextCRC (column “G”) → CRC value of the next segment;
src filename (column “H”) → name of the source file;
tgt filename (column “I”) → name of the target file;
segmentid (coulmn “J”) → ID of the segment.
The colored boxes are an indicator of how good was the fitting of the segment pair (calculated by the A.I. - it is estimated on the basis of the dictionaries used for the given language pair).
aligned.90+.xlsx
Its contents are the same as for the aligned.xlsx file besides the fact that it contains only the segments that have met several requirements:
The two segments cannot originate from two separate parts of the document.
The two segments must have the appropriate value of probability (>= 0.9).
The two segments must have the appropriate value of an anchorMatch factor (an internal phrase that is indicating an overall similarity of the sentences).
The difference in the character count for the segments cannot be significant (it is counted dynamically for each segment pair during the process).
Segments around them must also have appropriate match values.
Good to know!
Please, keep in mind that you cannot edit the source segments in the spreadsheet! They cannot be merged or edited in any way, as they represent the way in which the source text has been segmented.
On the other hand, you can edit the target segments to reflect the source segmentation if required. For example, you can merge multiple target segments into one cell.
In case of any issues with utilizing the aligning feature, we highly recommend getting familiar with our article on aligning constraints where we laid out the most common problems pertaining to this very module: Aligning - most common issues & troubleshooting.