It's simple math: The sheer volume of translatable website content multiplied by the number of languages needed to globalize can make a translation initiative out of reach. The truth is that machine translation (MT) offers access to otherwise unreadable information for global consumers.
While it’s true MT can't account for context and tone and often fails to grasp the nuances and ambiguities of language, there are several technology advancements improving the quality and efficiency of MT. Additionally, the MT engines offered by language service providers aren't the humble online translation tools of yore. They include complex rules engines and sophisticated algorithms, which are often also trainable and customizable based on industry terminology, resulting in much higher quality translation output.
Before dismissing MT, eBusiness leaders should consider that:
- Big data has made a big impact. Faster and more sophisticated algorithms allow for greater efficiency and accuracy in translation processes and workflows.MT relies on the coding and matching of languages — a vast and variable data set — thus, it has benefitted greatly from advances in analytics and data processing.
- Translation memory is boosting translation quality and efficiency. Translation memories are sophisticated language databases that store already translated content. By leveraging previously translated content, firms can reduce the volume of new content to translate plus improve consistency in translation outcomes.
- Translation management solutions (TMSs) are coming into focus. TMSs are essentially localization project management tools. They offer capabilities like workflow management and in-content translation editing. These tools can help enterprises manage the large number of content sources and translation services as well as help build a centralized translation memory repository (which can be especially useful when managing multiple language service providers).
Forrester published a report yesterday to help eBusiness leaders understand the role MT can play in eCommerce localization initiatives (client access required).
Have you seen MT leveraged artfully on any eCommerce sites lately?