Developments in Language Technologies Open Up New Opportunities for Companies
If you want to convince someone that machine translation (MT) is a technology best left well alone, you will find countless examples online to prove your point — just browse through one of the many websites dedicated to collecting content that has truly been “lost in translation” when processed by a machine translation engine. Using MT to translate snippets of text without any context is a sure-fire way to produce an inaccurate and, admittedly, often hilarious result. However, in reality, this type of fragmented content only accounts for a tiny fraction of the information that business customers submit for translation.
The world’s expectations of machine translation technology are another factor contributing to its somewhat undeserved poor reputation; perhaps we simply set the bar too high. Consider, for example, this quote from Dr. Thomas Goppel, then a member of state parliament and State Secretary of the Bavarian State Ministry of Sciences and the Arts, at the Machine Translation Summit II on August 16, 1989 in Munich: “When politicians see progress being made in the field of machine translation, they expect that they will have almost instant access to translated German versions of the massive collection of data and written records held by the European Union.”
Applications for Translation Solutions
In recent years, significant progress has been made in the field of machine translation, and free translation programs available online are now capable of producing acceptable or even high-quality results, depending on the type of source text that is fed into the engine. Traditional, word-for-word machine translation has now taken a back seat and is gradually being replaced by transfer-based, statistical or example-based methods.
These translation solutions are best suited to content that would otherwise remain untranslated in order to save money — for example, communications material for internal use within a global company. Employees need to be kept informed about what is going on in their organization, and today’s machine translation systems can serve as an internal communications tool to help companies do just that. Translating such content demonstrates to employees that the company values local languages and, most importantly, ensures that all information is communicated effectively.
How Can Translators Use This Would-Be Competition to Their Advantage?
Not all documents are suitable for translation by a machine translation engine — complex technical support requests, detailed EU documentation, pharmaceutical product information and assembly instructions for industrial machinery are just some examples of texts requiring human input. However, this does not mean that machine translation should be written off entirely when these kinds of documents need to be translated. Even professional translators, once seen as the potential unhappy victims of machine translation, are now embracing the technology to help them work more quickly and efficiently.
This new approach involves a translator running their source text through professional translation software, usually a translation memory-based system, which searches for matches in the existing translation memory. If the software identifies an identical or previously translated section of text in this memory, it can be reused in the new translation.
If no previous translation exists for any part of the text, the translation memory software can access external machine translation programs to provide the translator with suggestions, or a rough overview of a specialist field they may not be familiar with, in just a few seconds. Although the translator must check the suggestions provided by the software thoroughly, this approach can significantly reduce the amount of time a translator needs to produce a high-quality translation. As an added bonus for the translator, the basic principle of translation memory software continues to apply with the machine translation approach — the edited version of the machine translation is also added to the translation memory and will be available for reference and use in future translation projects.
The Moral of the Story
Combining translation memory and machine translation technologies can be an incredibly profitable move for companies that understand the opportunities this approach can open up, and know how to use the technology effectively.