What is translation technology?
Translation technologies are sets of software tools designed to process translation materials and help linguists in their everyday tasks. They are divided in three main subcategories:
Machine Translation (MT)
Translation tasks are performed by machines (computers) either on the basis of statistical models (MT engines execute translation tasks on the basis of accumulated translated materials) or neural models (MT engines are based on artificial intelligence). The computer-translated output is edited by professional human linguists through the process of postediting that may be more or less demanding depending on language combinations and the complexity of materials, as well as the volume of content.
Computer-Aided Translation (CAT)
Computer-aided or computer-assisted translation is performed by professional human translators who use specific CAT or productivity software tools to optimize their process and increase their output.
Providing a perfect combination of technological advantages and human expertise, CAT software packages are the staple tools of the language industry. CAT tools are essentially advanced text editors that break the source content into segments, and split the screen into source and target fields which in and of itself makes the translator’s job easier. However, they also include an array of advanced features that enable the optimization of the translation/localization process, enhance the quality of output and save time and resources. For this reason, they are also called productivity tools.
The most important features of productivity tools include:
- Translation Asset Management
- Advanced grammar and spell checkers
- Advanced source and target text search
- Concordance search.
Standard CAT tools include Across Language Server, SDL Trados Studio, SDL GroupShare, SDL Passolo, memoQ, Memsource Cloud, Wordfast, Translation Workspace and others, and they come both in forms of installed software and cloud solutions.
Quality Assurance (QA)
Quality assurance tools are used for various quality control checks during and after the translation/localization process. These tools use sophisticated algorithms to check spelling, consistency, general and project-specific style, code and layout integrity and more.
All productivity tools have built-in QA features, but there are also dedicated quality assurance tools such as Xbench and Verifika QA.
What is a translation asset?
We all know that information has value and the same holds true for translated information. This is why previously translated/localized and edited textual elements in a specific language pair are regarded as translation assets in the language industry – once translated/localized and approved, textual elements do not need to be translated again and no additional resources are spent. These elements that are created, managed and used with productivity tools include:
Translation Memories (TM)
Translation memories are segmented databases containing previously translated elements in a specific language pair that can be reused and recycled in further projects. Productivity software calculates the percentage of similarity between the new content for translation/localization and the existing segments that were previously translated, edited and proofread, and the linguist team is able to access this information, use it and adapt it where necessary. This percentage has a direct impact on costs associated with a translation/localization project and the time required for project completion, as the matching segments cost less and require less time for processing.
Translation memories are usually developed during the initial stages of a translation/localization project and they grow over time, progressively cutting localization costs and reducing the time required for project completion. However, translation memories require regular maintenance, i.e. cleaning for this very reason, as the original content may change and new terminology may be adopted.
In case when an approved translation of a document exists, but it was performed without productivity tools, translation memories can be produced through the process of alignment:
Source and target documents are broken into segments that are subsequently matched to produce a TM file that can be used for a project.
Termbases (TB)
Termbases or terminology bases (TB) are databases containing translations of specific terms in a specific language pair that provide assistance to the linguist team and assure lexical consistency throughout projects.
Termbases can be developed before the project, when specific terminology translations have been confirmed by all stakeholders (client, content producer, linguist), or during the project, as the terms are defined. They are particularly useful in the localization of medical devices, technical materials and software.
Glossaries
Unlike termbases, glossaries are monolingual documents explaining specific terminology in either source or target language. They provide further context to linguists and can be used for the development of terminology bases.
Benefits of Translation Technology
The primary purpose of all translation technology is the optimization and unification of the translation/localization process, as well as providing the technological infrastructure that facilitates work and full utilization of the expertise of professional human translators.
As we have already seen, translation memories, once developed, provide immediate price reduction (that varies depending on the source materials and the amount of matching segments, but may run up to 20% in the initial stages and it may only grow over time), but the long-term, more subtle benefits of the smart integration of translation technology are the ones that really make a difference and they include:
Human Knowledge with Digital Infrastructure
While it has a limited application, machine translation still does not yield satisfactory results that can be used for commercial purposes. All machine translations need to be postedited by professional linguists and this process is known to take more time and resources instead of less.
On the other hand, translation performed in productivity tools is performed by people, translation assets are checked and approved by people, specific terminology is developed in collaboration with the client, content producers, marketing managers, subject-field experts and all other stakeholders, eventually providing a perfect combination of human expertise, feel and creativity, and technological solutions.
Time Saving
Professional human linguists are able to produce more in less time. Productivity software, TMs, TBs and glossaries all reduce the valuable hours of research and translation, and enable linguists to perform their tasks in a timely manner, with technological infrastructure acting as a stylistic and lexical guide.
This eventually enables the timely release of a localized product/service, with all the necessary quality checks performed.
Consistent Quality Control
The use of translation technology itself represents real-time quality control, as linguists rely on previously proofread and quality-checked elements, and maintain the established style, terminology and quality used in previous translations.
For example, consistency and precision were major challenges in solutions that we developed for Mettler Toledo, and we were able to maintain a high level of quality throughout the project by using the benefits of translation technologies.
Brand Message Consistency
Translation assets enable the consistent use of a particular tone, style and intent of the brand in all translation/localization projects. This means that the specific features of a corporate message for a particular market/target group will remain intact even if the linguist team changes on future projects.
Code / Layout Integrity Preservation
Translation technology enables the preservation of features of the original content across translated/localized versions, regardless of whether the materials are intended for printing or online publishing.
Different solutions are developed for different purposes. For example, advanced cloud-based solutions for the localization of WordPress-powered websites enable full preservation of codes and other technical elements, save a lot of time and effort in advance and optimize complex multilingual localization projects.
Wrap-up
In a larger scheme of things, all these benefits eventually spell long-term cost/time savings and a leaner translation/localization process due to their preventive functions that, in addition to direct price reduction, provide consistency, quality control and preservation of the integrity of source materials.