With their flair for languages, translators are excellent usability testers for all kinds of texts. Translator queries frequently inform authors of anomalies in texts, therefore it makes perfect sense to use translator queries in order to improve the source text quality.

Other types of translator queries are terminological issues. With Smartquery you can include these queries in your termbase to increase the reach of your terminology, saving significant costs. Missing terminology costs time and money, both of which have a negative impact on translation projects for the client. However, translator queries so far were always considered a "necessary evil" that create work and slow down the translation process. Rarely were they seen as a potential to improve the source text and the translation quality!

Together with Austria's largest LSP, eurocom, Kaleidoscope decided to tackle this issue. In the process, we came up with clever answers to questions such as: How can you manage queries in a collaborative, web-based environment? How can you re-use queries as term candidates and also forward suggestions to improve the source texts directly to the authors? How can you create statistics on queries for a single client and perhaps discover and address specific problems systematically? How can you use translator queries to improve the translation quality in an effective way?

The Foundation

Because a scientific foundation for query management was lacking, the first step was to run a master thesis project with the University of Vienna to review and categorize large amounts of query data. Two large eurocom clients were used as the basis: 1,080 query forms in Excel format as well as the entire e-mail traffic about queries made up the corpus for analysis. The majority of all issues (about 80%) were terminology queries. The "other" queries were about the process itself, for instance issues with translator guidelines or with the source text itself.

At the same time, a pilot project concerning query workflow was conducted with another large client of eurocom and revealed some additional insights into the workflow design. The data from this pilot, together with the findings of the thesis project fed into the architecture of the Smartquery workflow solution.

Types of Queries

Not every query is the same. Based on the master thesis project we commissioned with the University of Vienna, Smartquery differentiates between three types of queries:

Terminological Queries

These can be logged by translators or project managers and are categorized as follows:

• Incomprehensible source term Translation suggestion
• Request for approval of target-language translation suggestion
• Abbreviation
• Proper name, label, software option, slogan
• Inconsistent source text
• Inconsistent target-language references

Some fields in the query might already be pre-filled if you are using Smartquery with an interface to a project management system. Also, you will only be able to choose the project which you are currently assigned to.

You can define these categories freely as well as all the fields for a query and the access rights.

Process Queries

Queries that do not relate to terminology may refer to the process itself, for example problems with the instructions, style guides, or the source text – more specifically incomplete sentences, typographical errors, deviation from terminology guidelines, technical problems, etc. This type of query can be submitted by all roles.

Feedback Loops

The feedback query category includes information that although does not pose a problem for translation, is still worthy of note.

Project managers and clients can initiate these feedback loops; however, a project manager has more options at their disposal than a client. Feedback loops started by a client can only contain a specific message. It will be visible to both project managers and translators. Feedback is always visible in the "My view" area for each user. In addition, users always have to acknowledge feedback by marking it as "read". The user who created the feedback loop can check the read status at any time.

If the feedback loop is initiated by the project manager, then it can be directed either to the translators or to clients (or both). In addition to free text, feedback loops can also contain specific queries relating to a project. To add queries, project managers can mark the relevant queries as "inform client feedback" or "inform translator feedback" when creating a feedback loop.

All queries can be assigned as many comments as you like and all actions are saved in real time. This enables all questions, answers, and comments to be fully tracked and documented throughout the process.

Managing Queries

If you want

• to organize translators' queries once and for all – in a comprehensible and traceable way;
• without relying on any Excel lists or other makeshift documents, which are never seen by everyone in the translation chain;
• and you also want to improve the quality of the translations and source texts at the same time;

then we have got the perfect solution for you: Smartquery!