STAR Transit
STAR Transit NXT is the most recent version of a long-established computer-aided translation (CAT) tool which is still in use in some Central European countries but seldom elsewhere. Users of the platform are often reluctant to migrate to other tools because of perceived difficulties in transferring the complex terminologies that have been created and maintained in the associated terminology management tool Term Star.
memoQ offers excellent integration with STAR translation products such as Transit NXT and earlier versions and with Term Star. Package files (PXF and PPF) from the STAR Transit environment can be imported easily into memoQ along with reference resources and terminologies bundled in the package.
Information on setting up STAR Transit packages in a memoQ online project can be found in this video: https://youtu.be/I-hwQlo9Goc
Information on setting up STAR Transit packages in a local memoQ project can be found in this video: https://youtu.be/AR9Ebwd7pQw
The second video also covers some workflow situations that may be relevant to online projects.
Terminologies provided separately, for example as MDB (Access database) files from Term Star can be exported in various ways to enable their import into a memoQ term base, Qterm or other terminology management tools.
STAR Transit projects at Linguaemundi[edit | edit source]
Most projects of this type are for the end customer Daimler and require the use of Daimler's specific, proprietary Term Star term bases. These term bases are not bundled in the package and must be attached separately (from the MDB file) to a project done with STAR Transit NXT, or they must exported via Term Star and imported into memoQ term bases, which are then attached to the package project in memoQ. There are currently two term bases of interest, one for "products" and another for "corporate identity.
If the projects are done in memoQ, the package project will generate a valid STAR Transit NXT "return package", which can be opened by the customer using a licensed copy of STAR Transit NXT.
Exporting terminology data from the MDB file for import to memoQ term bases[edit | edit source]
Term Star offers many options for exporting terminology data to use in other tools. However, in the case of very complex terminology collections such as Daimler's term bases, the exports will inevitably involve some compromises and data loss. The distributed MDB files typically contain many languages other than those of interest for a particular project or language pair, and there are also a lot of synonyms, forbidden terms and other information which may get lost or become unclear after export. When in doubt, consult an original entry in Term Star to see any relevant metadata or other information.
For purposes of Linguaemundi's projects in memoQ, the best export format to use from Term Star appears to be TBX.
The TBX exports contain many languages of no interest, which will need to be removed from the memoQ term bases after the data are imported. Moreover, TBX imports to a memoQ term base override the default term properties of the term base, an unfortunate thing which does not happen with most other import formats. This can be considered a bug in current versions of memoQ.
Importing Daimler TBX exports to memoQ term bases[edit | edit source]
It is recommended to create two new memoQ term bases, one for Daimler Products and another for Daimler Corporate identity. Some system of versioning, such as the addition of a date to the name of the term base, should be used so that people will be aware of the version status of the information.
Initially, the term base will need to have only two languages specified (preferably language variants used in Linguaemundi projects); the TBX import will automatically add all the languages contained in the file. Unwanted languages can then be removed by unchecking the boxes for them in the Properties dialog of the term base. Consult previous term bases for what languages to retain or ask the quality manager for guidance in this matter. If languages are not currently part of the work with the client they may still have reference value, because they will be included in the display of the term hit in memoQ and may be useful suggestions for possible terms that are missing in the target language of actual interest.
After the TBX file has been imported to a memoQ term base and the unwanted languages have been removed, the next step is to change the match properties for all the relevant languages to "Fuzzy". Although this may lead to false positives, it will ensure the best coverage of possible terms regardless of typos, misspelling, grammatical structure changes, etc. In cases where these false positives become a nuisance, the terms should be entered redundantly in Linguaemundi's own Daimler term bases, where the properties can be fine-tuned for QA purposes.
Changing the term match properties can be done in memoQ's term base editor by a bulk (range) selection process in each language set. If a range (of all) of the terms is selected and the match field is set to Fuzzy, the match property will be changed for all the selected entries. This can be a slow operation in large term bases, and it is usually more efficient to do this work in a local term base which is later published to the server after all adjustments are completed. Due to the nature of the entries in Daimler's term data, a setting other than fuzzy matching for terms will very often result in the terms not being found, so this is very important.
The change of term properties can also be done by exporting the data from the memoQ term bases, including the Entry ID and Term Info, importing the Term Info column for "translation" in memoQ, then exporting the modified CSV or Excel data and re-importing it to the respective term base with the setting to update the term base using the Entry ID. This may prove to be much faster, and it would allow the work to be done in a single pass for all languages by importing only the Term Info columns of an Excel file, for example.
Finally, it is important to preserve the customer's data as received. So these term bases should be made read-only before they are used in projects. Any changes or adjustments to term entries should be done by putting the terms into Linguaemundi's own Daimler term bases or personal, local term bases, not by editing the collections received from Daimler.