Concordance search


XTM Cloud contains a separate Concordance module:

Concordance as a separate module in the XTM UI

It is also accessible directly in XTM Workbench as a separate dedicated docked panel:

Concordance in XTM Workbench

Apart from that, you can also search for a particular source/target string in a segment directly from the segment context menu in XTM Workbench, once you right-click on that segment.

Concordance from the segment context menu in XTM Workbench

To see a step-by-step guideline to activating and using Concordance, click here.

How does Concordance search actually work?

You might wonder about the difference between the Concordance search result and the TM search results. This is the explanation.

Although Concordance draws its data from XTM Cloud translation memory, it is a completely separate module, and it works in a different way.

Although Concordance draws its data from XTM Cloud translation memory, it is a completely separate module, and it works in a different way.

Its sole purpose is to give the most accurate results and it does so in the following way:

  1. It searches for a particular expression across the TM database.

  2. It then limits the search results to 50 only.

  3. It rejects all the duplicates among these 50 results.

  4. It displays the best-matched results in the panel, starting from the top.

  5. You can then sort the results by selecting one of the criteria in the Sort by section. When these are used, the results are first sorted by the selected criterion and secondly by the score. The default criterion is Score:

In other words, Concordance displays 50 (source or target) results from the TM (depending on your search criteria) and removes duplicates amongst the displayed phrases.

For example, if you were to have 50 counts of TM:

(English) Alice has a cat. <--> (Polish) Alicja ma kota.

and the 51st TM entry was:

(English) Alice has a cat <--> (Polish) Alicja ma psa.

Keep in mind that this last unique entry might not then be taken into account by the Concordance search.

Searching based on matching type

The mechanism described above applies to any search criteria, but you can still set the type of segment to be chosen as a search basis for the requested expression by selecting one of three match types:

  • Broad match → finds segments with any of the words in the search phrase along with other words or characters.

By default, the broad match search engine is used. It applies stemmed forms such as Load balanc, Load balance or Load balancing in the search results for the searched-for phrase: "Load balancing".

  • Phrase match → finds segments with all the words in the search phrase, but in any sequence.

  • Exact match → finds segments that only contain the words in the search phrase and in the sequence strictly defined in the search field.

Considering the above, more accurate results might be displayed in the Concordance search by using "Phrase match" or "Exact match" options. The most TM-alike search would be the "Phrase match" option.

You can also choose to perform a Concordance search with source and target language words, or phrases, either in the whole set of TM databases connected to the project or just in the Approved TM. Not approved status is displayed as an icon. If you are not sure what the status of the TM search result is, left-click the Metadata column to open a popup with additional details about the segment.

LSP case: the XTM Cloud UI vs. XTM Workbench

There is a significant difference between the Concordance function in the XTM Cloud UI and the one in XTM Workbench as far as subcontracting is concerned.

Concordance in the XTM Cloud UI only permits searches in the local translation memory, which is the subcontractor's own TM. In other words, if a particular LSP does have access to the Contractor's TM, they will not find it there.

On the other hand, the Concordance function in XTM Workbench can be used to perform searches of this kind across client translation memory data.