With the rise and ever-evolving presence of computing technology, the translation industry has found itself constantly changing as it integrates new technological marvels into its world with each passing year.
The development of translation technology tools, also known as a translation management system, has been invaluable. With the rapid rise of globalisation and the global interconnectedness brought about by the internet, translators are finding that that they have an increasing demand placed upon them in the form of increasing work loads. As such, by applying such technology and translation tools to the development of translation, language service providers and translation agencies have been able to boost the productivity and translation workflow of their professional translators and offer quick turnaround times for their clients’ projects. As a result, the integration of translation tools has become pivotal in the attempt to increase project output as well as strengthen market competitiveness.
Translation Tools
It’s safe to say that nowadays, the means of translation has become more streamlined and less time consuming. As a society, we’ve moved away from relying solely upon physical dictionaries (having to emphasis the “physical” element is already an indication as to how far we’ve come), hand-written glossaries, and our individual knowledge of a particular pair of languages.
The translation tools that are currently available tend to fall under the umbrella of Computer Assisted Translation (CAT). CAT tools are simply the utilisation of computer hardware and software by human translators as a means to aid and bolster the translation process. Computer Assisted Translation tools come in a variety of different categories, including, but certainly not limited to Terminology Databases, Translation Memory, and Language Search Engines, allowing users to access terabytes of data for the purpose of translation at the push of a button.
However, we’ve progressed as a society, and the measurements of what makes translators good within their field has shifted. The translator’s see-saw of “quality and accuracy” vs. “speed and efficiency” has certainly tilted more towards the latter as technologies become more and more prominent and their use in the practice of translation becomes increasingly apparent. Whilst professional translators aren’t necessarily losing their jobs to platforms such as Google Translate and similar online web dictionaries, as people tend to use these applications as a means to get a general idea of what a text is about, people do utilise translation agencies or freelancer translators that have integrated translation technology tools for their official translation pieces. The integration of such systems has become a necessity for translation services and freelance translators to remain competitive within a somewhat bloated market. There is a wide range of possible systems and resources available for a business looking to manage their translation processes and offer a solution to company costs and quality assurance for their project managers. These range from various types of software for translation memories, search engines, and terminology management, to the more standard tools like spelling and grammar checkers, pre-existing terminology bases, and online dictionaries.
Common CAT Tools
Whilst there are plenty of platforms, applications, and software available for the task of translating and localization, as an agency with years of prior experience within the translation industry, we have highlighted two major applications that are very popular amongst translators and agencies alike:
MemoQ
MemoQ is one of the most popular and widely used CAT tools used by many professional translators around the globe. Not only is it user-friendly and safe (and accessible on any web browser), but MemoQ has a well-earned reputation within the industry due to its streamlined workflow and localisation processes designed to facilitate the translation process, with the platform providing Translation Memory, Terminology Databases, and Machine Translation.
As a tool, the platform offers a range of invaluable functions and features for all projects and the development of quality content, including the ability to reuse previous translations; improve writing quality; and check work consistency and the use of correct terminology.
SDL Trados
SDL Trados, available from SDL plc, a leading provider of customer experience cloud solutions and services, provides a comprehensive suite of utilities for independent translators, agencies, corporate translation departments, localization providers and academic bodies. SDL Trados guides the user through all stages of the translation process, from big-picture project administration and glossary management to proofreading, editing, and building a Translation Memory.
The core of SDL Trados offers its users the ability to edit or review translations; manage projects; organize terminology; and connect to Machine Translation.
However, whilst SDL Trados is a fantastic way of improving translation efficiency, it can be a little expensive for users to integrate into their pre-existing systems and services. As such, it is certainly a great choice for medium and large-scale agencies.
Machine Translation
Machine Translation (MT) (not to be confused with CAT) is a process of translating text from one language to another in real time using computer software and without any human involvement.
Benefits to Machine Translation
Overall, MT saves a significant amount of time by being able to translate entire text in the space of a few seconds. Plus, it is a cost-effective practice due to the quick turnaround time and lack of human involvement. However, perhaps one of the most exciting aspects of using such technology is its learning capabilities. Just like how human translators learn and get more effective with each project undertaken, MT performs the same marvel but without the problems brought about by the human brain (such as poor memory and self-doubt). Consequently, its ability to memorise key terms and reuse them again in the future whenever necessary makes MT an invaluable and impressive tool.
Drawbacks to Machine Translation
On a basic level, MT undergoes a process of word-for-word mechanical substitution. For professional translators, translating word-for-word is a massive no go area as in doing so can affect the accuracy of the overall text. As such, utilising such technologies can affect the quality of the final text as there is no recognition of linguistic elements pertaining to language such as whole phrases, idioms, and slang. Plus, not all words in one language have equivalent words in another, with many words having more than one meaning. As a result, it is important to keep in mind that when using such a tool, human translators should always post-edit any work undertaken by MT.
Machine Translation Types
Whilst there are a wide range of MT types, we have created a compilation of the 3 main types of MT that are worth taking note of when diving into translation technology:
Rule-Based (RBMT)
RBMT (or the “classical approach” of MT) is a tool that uses grammar and language rules and is based on linguistic information about source and target languages retrieved from (unilingual, bilingual, or multilingual) dictionaries and grammars covering the main semantic, morphological, and syntactic regularities of each language respectively.
Having input sentences (in some source language), an RBMT system generates them to output sentences (in some target language) based on morphological, syntactic, and semantic analysis of both the source and the target languages involved in a concrete translation task.
Statistical (SMT)
SMT tries to generate translations using statistical methods based on bilingual text corpora, and is a method undertaken by web translation tools such as Google Translate. Whilst this tool doesn’t rely on linguistic rules and words, instead learning how to translate through the analysis and access of large amounts of pre-existing human translations, accurate and high-quality translation pieces are only developed when there is a healthy foundation of pre-existing data and corpora resources available.
SMT’s biggest downfall includes it being dependent upon huge amounts of parallel texts; its problems with morphology-rich languages (especially with translating into such languages); and its inability to correct singleton errors.
Neural (NMT)
NMT is an approach to MT that uses an artificial neural network to predict the likelihood of a sequence of words, typically modelling entire sentences in a single integrated model.
This method is becoming more and more popular as it provides better results with language pairs, and it only requires a fraction of the memory needed by traditional SMT models.
Translation Management System
Translation Management Systems (TMS) are designed to manage the localization and translation of language assets and is very beneficial in the management of scaling translation by automating all repeatable and non-essential work, and leaving the creative element of translation, as well as all editing and revisions for a human being. As such, by aiding in the scaling process, these management systems can help companies organise and manage their translated assets, allowing all those involved, from project managers to translators to collaborate more effectively.
One of the most exciting business aspects of TMS is their ability to go further than simply translating a given text, as they are also capable of maintaining brand vision and identity by ensuring that the final content is contextualised.
Content Management Systems
Content Management Systems (CMS) play an important role in the management of translation projects, with its main selling point being the ability to have multiple contributors create, edit, and publish content. Its user-friendly approach means that the integration of such systems allows for a quick turnaround on project work as well as being highly cost effective, SEO-friendly, and offering a range of differing features depending on the users’ needs. Examples of CMS include WordPress, Squarespace, and Magento.
Whilst CMS is a fantastic way for translators to streamline their project work, it can be tricky for users to export data or move away to another platform. Plus, users must be cautious of any hidden costs (such as with implementation, customisation, support, and training) as well as being mindful of any upgrades and patches needed to ensure that all translation software is continuously secure.
Translation Memory
Put simply, Translation Memory (TM) is a dictionary of previously translated sentences and is perhaps the translation tool most commonly used by translators in their day-to-day work. It is also typically used in combination with a CAT tool.
TM is used as a means to suggest identical or similar texts that have previously been translated in order to translate new texts and content more consistently and effectively. When the software recognises that a new segment is similar to a segment already translated, it will suggest that the translator reuse it. Not only does this make for a smoother editing experience, but it also speeds up the translator’s workflow.
How Does It Work?
TM works at the sentence level. As previously outlined, when a translation is being carried out using CAT applications, the sentence being translated is compared to other sentences within the memory databank. If a similar sentence is found, the CAT tool will inform the translator and provide the option to i) use the recommended translation ii) replace it with a new translation iii) modify the reference provided by the TM.
Benefits and Drawbacks
TM is not only cost efficient, but it’s also a very useful tool that is most beneficial when the translation project’s content contains a highly repetitive text (such as legal contracts) and also when translating incremental changes in a previously worked on document.
However, as already hinted upon, TM is much less useful when it comes to creative or literary texts due to the lack of repetition and the unlikelihood that the translator has translated the same sentence previously. Yet, with that being said, some translators find translation memories valuable even for content that is not repetitive. This is because the memory management system can be invaluable in providing a reference to help determine the appropriate usage of particular terms and can offer a more streamlined editing process.
Connect With Us
The world of translation can be a daunting and overwhelming place for the novice translator. With years of industry experience, Espresso Translations certainly knows how to safely navigate its way through the field of translation technology applications.
We are always at hand should you want to know more about translation technologies, and as a company, we are always happy to help and connect with those who are interesting in learning how such applications can be useful for their translation projects.
Feel free to get in touch with us today via our website to see how we can assist you and help get the most out of your translations.
