Maintaining linguistic resources is something users can do online and offline. Depending on the volume of work to be done and of the output expected, users performing these actions will decide to work directly in the online editor or export the resource and run some extensive clean-up with dedicated tools.
| Table of Contents |
|---|
| Info |
|---|
New See this article on our Knowledge Base to learn how to enable and use the feature: Monitor the quality of your Resources with the Maintenance and Statistics Dashboard |
Activities to clean-up the resources
Maintaining linguistic resources is something users can do online and offline. Depending on the volume of work to be done and of the output expected, users performing these actions will decide to work directly in the online editor or export the resource and run some extensive clean-up with dedicated tools.
| Table of Contents |
|---|
Cleanup activities via the CAT Editor
Phase 1: filterMaintenance activities in the CAT Editor
Step 1: Filter relevant segments in the Editor
| title | Use the search bar with one or several languages |
|---|
using the Search Bar
The power search bar on the top of the
Editor will allow you to filter on specific criteria you may need to browse contents through pairs of segments.
In the language menu on top, you could
filter up to 2 languages
.
Additionally, you could
filter segments based on the values for each language.
This means you could search for segments which contain specific texts for each language (where a text is translated in a specific way, or where the text is present in both, source and target segments).
Filter options
It is possible to combine several categories to apply a custom filter. Once you enter the values you expect to apply, the categories are enabled and their field name is highlighted.
The relevant filter categories and their purpose are listed below:
Table of Contents maxLevel 5 minLevel 5
Option 1: The text contains: filter on segment text
The text contains filter is equipped with this search operator (AND/OR), among other search possibilities such as phrase match, case-sensitive search or even Regex /wildcard search.
For example, in
an EN-GB to
CA-CA job, you could
apply a filter where the source segment has
a specific translation (filter on segments where "university" is translated as "college") or where a text is identical in both languages
.
Option 2: Bookmarks: filter
on pinned segments
The bookmarks filter is equipped with this search operator (AND/OR), which allows filtering on specific bookmark value(s) for both or each language(s)
Step 2:
Apply batch action
Once you have
decided on the filters and search criteria, you may want to update the segments with some specific property or value. The fastest way to update a group of segments is to use the batch actions menu.
Actions that may be relevant for resource maintenance activities are:
Language actions menu:
Clean all segment information: removes all contents for the given target language (this includes translations, bookmark, status, labels and any other related properties)
Set custom QA message: if you want to leave a custom "error" message which could help you do further research on what needs to be done for those segments.
Language-specific label: if you use labels for maintenance purposes. See labels (category: resources).
Find repetitions: This action lets you find multiple occurrences of identical source texts and translations.
Example of use cases
:
Find source text repetitions.
Find redundant segments that have both identical source text and translation.
Find identical source texts that have very similar translations.