Technical terms

Technical terms

WHAT DO WE MEAN WHEN WE SAY … ?

Abbreviations and acronyms in the translation industry generally originate in the English language and are not necessarily self-explanatory. Here we explain of some of the most frequently used abbreviations.

ICR (In Country Review)

A review of the translation by a native speaker living in the target group region.

This is often done by the sales or marketing manager of the foreign branch of a company, who has excellent knowledge of the product, services, and form of address in the target market. ICR feedback improves the collaboration with the translator, improves quality over the medium term, and prevents extensive post-processing.

QUERY

Term used for a question during the translation project.

Depending on the quality of the text, specialization of the topic, or pursued objective, it is possible that questions arise during the translation work. Complex translation projects for which many files must be translated into several languages require sophisticated query management, which the project manager uses to control the flow of information. Query management is closely tied to terminology and memory management and plays a key role in quality assurance.

CAT (Computer Assisted Translation)

Computer assisted translation is done by humans with the help of computer programs.

Unlike MT, these programs do not translate automatically but assist the translator in the translation process by providing stored translation units (translation memory) and terminology databases. CAT tools help optimize quality, delivery times, and costs.

MT (Machine Translation)

Translation done by machines.

In machine translation, computer software translates a text from one natural language into another natural language. Complex rules are developed because word-for-word translations rarely provide comprehensible results. Acceptable results can now be achieved for highly standardized texts (depending on the language combination). All such translations must be post-processed by a human translator.

ML (Machine Learning)

Machines that learn.

MT in combination with auto-adaptive ML offers a new paradigm. Such systems learn from the experience, the intelligence, and the findings of their human users. They increase productivity, offer suggestions, and increase accuracy over time. The Californian company Lilt, a start-up of Stanford University and partner of FaustTranslations.com, is working on this innovative technology, which will define the translation industry in the future.

STYLE GUIDE

Design guideline that describes how a translation should be laid out.

The style guide creates consistency among different texts and thus helps create or secure the corporate identity. Wording and presentation is standardized.

The style guide may, for example, specify grammar rules, fonts and font size to be used, as well as the format for captions, etc. Some style guides are more than 100 pages of instructions and examples (e.g. Microsoft). In practice, a much smaller style guide usually is sufficient.

TM (Translation Memory)

All translated units are stored in the TM.

CAT tools divide the source text into segments. Most commonly this is one grammatical sentence. The translated segment is stored in a database and can be used for future translations. A well-maintained TM can offer significant time and cost savings.

MATCH (Translation Matches)

A “match” is the level of similarity of a segment in the TM compared to a new segment.

CAT tools make a translator’s work easier by comparing new texts to existing translations in the TM. Segments that are a perfect match are called “100% matches”. There are also 101% matches (if the segment before or after is also identical) and 102% matches (if the segments before and after also perfectly match a previous translation) that provide an indicator of congruence within the context.

Lower matches (e.g. 78%) indicate a certain difference between the stored translation and the new source text. We do not charge for 100% and higher matches.

REPETITION

Segments that repeat within a file/project.

Exact repetitions of a text segment within a file or a project with several files are called repetitions, even if they occur in different files within the same project. CAT tools can recognize repetitions and include them in the analysis. Repetitions must be translated only once and therefore offer potential cost advantages if texts are created with identical text modules. We do not charge for repetitions.

ALIGNMENT

Creating a TM from existing translations.

If translations already exist in a digital format (source text and target text) and if they are to be used for future translation projects, then an alignment can transfer them into the TM. In this process, the source text and the target text are placed next to each other so that the relevant segments can be allocated, reviewed, and stored. This makes previous translations immediately available for new projects and they can be reused cost-effectively (see MATCHabove).

TB (Termbase)

Technical terminology glossary.

CAT tools manage a TM as well as a TB with relevant, customer-specific, technical terms. Using the right terminology is key for a successful translation project. In the translation industry, wrong terminology and inconsistencies cause nearly half of all post-processing work, which affects costs as well as delivery time. Our TB management in SmartCAT is state-of-the-art. We use the customer’s existing terminology and then continuously update and manage the TB. Throughout this process, the customer has direct access to the terminology in our system.

QA tools (Quality Assurance)

Integrated tools to ensure the quality of the translation.

QA tools offer comprehensive options to review translations for errors. The translator is informed of possible errors already while working on the translation. Today’s functions go far beyond a simple spell check: they consider grammatical rules in the target language as well as numbers and number formats, date formats, punctuation, formatting, prohibited terms, trademarks and brand names, number of characters in one segment, etc. We use lexiQA for additional quality assurance because it is one of the most cutting-edge QA tools. The project manager can use a detailed report to ensure that the translation has no defined errors.

ANALYSIS (CAT supported)

The project manager uses the analysis to compare the text to be translated with existing translations.

CAT tools store the translated segments in a database so that they can be reused for future translations. The analysis calculates the number of already existing translation units and the number of repetitions within the file. The quote for the new translation and invoicing is based on this analysis. Intelligent creation of source documents can achieve significant savings in terms of delivery time and cost.
We never charge for repetitions and 100% or higher matches.

CONTEXT

Context plays a key role when working with CAT tools. The percentage of a match indicates the similarity of a segment in the TM to a new segment.

CAT tools are intrinsically “stupid”. They store translations as an identical concordance of a source language segment with the target language segment. They do not recognize the context of the sentence. This difficulty can chiefly be solved by reviewing whether the segment just before or just after the relevant segment is also identical (101% match) or if both are identical (102% match).

TERMINOLOGY

Consistent industry- or customer-specific technical terminology.

The importance of consistent terminology cannot be emphasized enough. Nearly half of all translation errors and half of all post-processing work is caused by wrong or inconsistent terminology. These are technical terms the use of which in a translation project is obligatory. A glossary forms the basis of every translation. It is either provided by the customer or is created by the translator and then coordinated with the customer. With collaborative workflows, this can now be done relatively easily and quickly. The initial (and admittedly somewhat tedious) effort quickly pays off because it reduces error sources, makes queries unnecessary, and achieves an optimum quality of the translation.

COLLABORATIVE WORKFLOWS

Cloud-based collaboration within the translation environment.

Most recent technical developments have changed the way in which we collaborate for the long-term. Tasks that used to be done in sequence or that required time-consuming communication are now integrated and overlap.

Within our SmartCAT translation environment, everyone participating in the translation project – translators, editors, project managers, terminologists – can simultaneously access all resources like TMs, TBs, dictionaries, etc., and everyone can communicate within this environment. In real time. The customer’s employees can easily be integrated. This saves time, improves quality, and turns collaboration into a smooth and pleasant process.

BACK TRANSLATION (BT)

Translating back into the source language.

A back translation is an additional step in assessing the quality of a translation by having an independent translator, who is not familiar with the original text, translate the translation back to the source language. A comparison of both texts then shows possible comprehension or translation errors of the first translation, which will have to be corrected.

This is a complex (and thus costly) step in quality assuranceand it is generally used only for sensitive content (e.g. pharmacology, aeronautics) and cross-cultural marketing contexts.

CLOUD COMPUTING

Internet-based use of virtual resources.

Cloud computing is basically a technological relocation of the location where data is processed and stored: from local PCs to remote data servers, through an internet connection. Modern technologies have advanced to the point where a time delay can no longer be discerned. Extensive resources, maximum processing power, and networking with other users are clear advantages that enable collaborative workflows.

Concerns regarding data safety, which are frequently expressed, can be refuted because our data is safe, just like data stored on local or cloud-based servers at public authorities, insurances, banks, credit card companies, e-shops, etc.

« »