Linguists familiar with CAT tools will feel at home with the Smartcat editor. Many of the functionalities (and even the hotkeys) are similar to what can be found in other applications although with a stronger emphasis on communication and collaboration. One major difference would be in the multi-user interaction where linguists at all stages can be working on the same document at the same time. Smartcat implements a locking mechanism that enforces the translation workflow while at the same time facilitating collaboration, saving time and money in most projects.
The interface is made up of five components:
- The toolbar where the users will find buttons for most of the editing functions.
- The editing area with four columns for source and target text as well as informational data.
- The CAT pane with proposals from the TM, MT and terminology.
- A series of tabs for communication, QA and revisions.
- A CAT information pane
From left to right, the buttons are:
- Confirm segments — this button is used to confirm segments although linguists will generally use the Ctrl-Enter hotkey. The down arrow on the button can be used to confirm all segments at once if the project manager activated this option for the language team. It can be useful for later stages of the process or a client review where the user might only need to correct a few segments and doesn't want to have to confirm segments manually one at a time.
- Go to the next unconfirmed segment (F9)
- Go to a segments by number (Ctrl-G) — allows the user to go to a segment by specifying it's number.
- Undo the last action (Ctrl-Z)
- Redo the last action (Ctrl-Y)
- Copy the source text to the target segments (Ctrl-Insert) — it is also possible to apply it to the entire document by clicking on the down arrow to display that option.
- Insert a special character (Ctrl-Shift-I) — Smartcat will display a table of special characters that can be inserted in the text. The user can add custom special characters to that table.
- Cycle through cases (Shift-F3)
- Insert tag (F8) — tags will be inserted in the order that they appear in the source document.
- Upload an image (Alt-Insert) — this will display a dialog box where the user will be able to add a link to an image or upload an image file for a specific segment. An icon will be displayed in the source column to indicate that an image is associated with that segment. This option can be useful when translating software strings and a screenshot can help with context.
- Segment length limit — this option is used to specify the maximum length of a specific segment. This is also useful when translating software strings.
- Add a new term (Ctrl-E) — used to add terms to a glossary associated with the account. Linguists might only the right to suggest terms.
- Return the segment to the previous stage — this option is only available to the project manager. It allows the PM to revert a segment to a previous stage of the process. For example, from editing to translation. This can be used when many mistakes make it necessary for the previous linguist to review many segments.
- Open the custom dictionary associated with the spellchecker.
- Concordance search (Ctrl-K)
- Merge segments
- Split segments
Some of these options might be invisible depending on user rights and choices made by the project manager.
The editing area is similar to what can be found in other tools with one read-only column for the source text and one column for the editable target text. In addition, there is one column to the left for segment numbering and repetition indication:
In this case, segment 3 and 4 are repetitions in the project as indicated by the down arrow. In segment 4, the user has indicated that this segment should not be auto-propagated with repetition by clicking on the arrow. It is indicated by a strike over the said arrow.
And an additional column to the right for segments status information and QA notifications:
For example, in the screenshot above, the top segment is in translation and suffers from a critical error (red triangle) while the second segment is already at the editing stage and has a QA warning. For both segment, the content came from a 102% match from the TM.
Comments to segments are indicated by a text bubble in the source segment:
Unread comments have an orange bubble which turns gray upon reading.
The top of the editing area has the text search and filter options:
The Source and Target fields can be populated by selecting text in the respective columns and using Ctrl-F. The text can also be typed in manually. By clicking on the Replace button () (or Ctrl-H), the user opens the Replace with field along with Replace options:
The Case check box () ensure that capitalization is taken into account for searches. The filter button at the extreme right shows how many segments match the criteria (2 in the example above). The Clear all filters button () simply resets any searches or filter active.
The Filter button can also be used to start some custom searches. The user can filter segments by status, by source text criteria such as repetitions or numbers, to show segments with comments or error and by the origin of the last input:
Each criterion has different options that can be accessed by clicking on the filter name.
The CAT pane displays matches from the translation memory, machine translation and the terminology database. By hovering over the origin of a match, the user can tell the exact source — TM and glossary name or MT engine:
In this example, the CAT pane displays one TM match (TM), 2 glossary entries (TB) and one MT match from the DeepL MT engine. The number on the left are used by the linguist to insert the text in the target field using the Ctrl key. In this case, Ctrl-6 would insert the TM match while Ctrl-7 would insert the glossary entry at the cursor location.
The CAT info pane is used to display the differences between the new source segment and any matches found in the translation memory as well as perform concordance searches:
In this example, Smartcat has found a 90% fuzzy match and the CAT info pane displays the differences, highlighting the addition with a green background and deletions with a reddish one. The pane also displays metadata for the match as the name of the database, the date of creation of the unit and the author.
Concordance searches will allow the user to search throughout the associated translation memories for a term in the source or target text:
The editor also includes a series of tab that are used by team members to communicate but also to display QA errors and segments revisions:
The Revisions tab will display all the changes that have been made by team members at any time during the translation process. The same scheme used in the CAT info pane is used in this tab — addition with a green background and deletions with a reddish one.
The tab will also include information on the name of the user who modified the text and time and date of the action.
The QA check tab will list information about the QA errors in the current segment. Linguists will also have the option to ignore some errors when for example, dealing with false positives.
The Comments to Segments and Comments to Document tabs are used for real-time communication between the team members. For example, project managers can leave instructions for the team in the Comments to Document tab and linguists can exchange ideas at the segment level.
The Settings wheel (to right corner of the interface) opens the Settings dialog box which contains options that can facilitate the work of the linguists.
The linguist can ask Smartcat to automatically insert text in the target field upon reaching a new segment. Rules are implemented one at a time and this will speed up the translation process as the linguist will not need to consult the CAT pane in most cases as the best option will be inserted in most cases.
This tab allows the user to choose how repetitions are being propagated in the document. Here is a quick explanation of some option:
Propagate to segments with different case — There could be times when the text in a segment is a repetition of another segment, but the words are capitalized differently. If this option is checked, Smartcat will propagate the text to segments regardless of the capitalization but with two possible options:
Use the original case — The capitalization of the segment will be maintained
Use the match case — The capitalization of the segment will be changed to match the original segment being propagated.
Propagate to confirmed segments — If this option is checked, the content of segments that have already been confirmed will be overwritten by the content of the segment being propagated.
Confirm auto-propagated segments — Segments that are being translation using auto-propagation will be confirmed automatically to match the current workflow stage.
Note: If the project manager set repetitions settings at the project level, linguists will not be able to modify these settings.
These options let the user decide what happens once a segment has been confirmed. In most cases, going to the next unconfirmed segment will be the preferred option as Smartcat will then skip any segments that would not need to be revised. In other cases, like for example a customer review, the reviewer might decide to review every segment and confirm them in batches at the end.
If given permission by the project manager, linguists will be able to download files from the editor:
Being able to download the original file could be helpful for context of the source text, while the translated file could provide the same context for the translation.
Smartcat offers three layout options for the user to choose. Each one of them present a different window layout that might be more adapted to the way the linguist is working. Users can experiment with these three options and choose what works best for them.
The up arrows at the end of the bar are used to hide the button toolbar to increase the size of the editing area and display more segments.