Translation is not enough.
Sometimes your overt intent in developing eLearning is to communicate a sequence of steps in a process, standards, or other facts. At other times your intent may be to persuade the consumers of eLearning to act or to make a judgment call. Sometimes you are hoping to do both.
The problem is that research has shown that every reader will prefer to see or hear the content in their native language. Nonetheless, a literal translation of the facts and the steps or the reason and the standards for the action will probably lose something in the translation. The "something" that gets lost is the nuance that only a culturally accurate product can provide. Getting the message across to readers in different countries, cultures, and experiences requires localization.
Saving the lost meaning
AI alone, even with its shortcomings, can produce minimally acceptable literal translations. Some products can automatically handle changes involving weights and measures, currencies, and related "tweaks" that help credibility. AI can localize multimedia content and it can automate workflows. It can facilitate consistent delivery of the literal message across language barriers. Is that good enough for a very complex eLearning project with lots of technical language?
Using a conversational AI product is one approach to getting at least some of this necessary refinement into your translation. It is still not enough. Even the most linguistically accurate translation achieved by AI can still sound "off" or wooden. This can make the transmitted message less credible, unfortunately. AI alone does not do humor effectively, for example.
Localization software
To minimize or eliminate the damage to the content, the translation must be localized for the place where it will be consumed. Currently, the ideal answer to is to include human editors on the translation team. That’s especially true when the content is intended to support expertise in healthcare, research, retail, financial services, telecommunications, and government.
Adding editors is not the only way to localize translation or even the best way. Many translation apps and platforms include localization features. In fact, when you search software review sites, the reviews themselves will include comments about localization features.
Localization apps should handle content for desktop, web, and mobile use. Localization software should have the ability to provide alternative translations and support rephrasing translations. Some software also has the ability to shorten translations in order to fit them to the space available. Other applications can build translation glossaries to prevent inconsistencies in the translation.
Software review sites: translation and localization
Capterra: After you open the link, scroll down the page to find a more comprehensive listing of translation and localization software. Read the actual reviews carefully. Some of them are a bit nuanced.
TrustRadius: TrustRadius does not have a separate section for localization—the translation reviews include comments that tell whether individual products cover localization. I also recommend reading the article at the very end of the reviews on computer-assisted software. It’s titled “Learn More About Computer-Assisted Translation Software 2023.”