Building a multilingual knowledge base for international customer

Learn how to evaluate, create and manage a multilingual knowledge base to support international users and customers in multiple languages.

What exactly is a multilingual knowledge base?

According to Wikipedia, multilingual relates to multilingualism and indicates the use of more than one language. So a multi lingual knowledge base holds content of at least two languages or even more.

What makes a good multilingual knowledge base?

From the end users perspective, all content of the knowledge base should be available in his or her language. It is expected to see all necessary articles and resources (as images and pdf files) in the requested language in the latest version. If the user accesses the documentation portal, it should be automatically detect the language and show the content in this locale.

If a new feature is added to the product, the documentation should be available not only in the primary language but also all other supported languages.

For the editor of the knowledge base, it should be easy to add new translations in multiple languages and see at a glance which articles and content are not yet translated yet.

3 quick steps to create a multi-language knowledge base

Creating and initializing the localization of a multilingual knowledge base depends mainly on the used software. Normally it can be sum up with the following three steps:

1. Initialize the knowledge base portal: the primary language is defined (mostly main language of company or of the audience as e.g. English) and additional, supported foreign languages are added.
2. Create folder structure and write the articles in the main language
3. For each locale, translate all content elements from step 2.

Of course, this just a short summary. The translation of the knowledge base articles can be done in-house, e.g. from a member of the support team, or a product manager that speaks the foreign language.

How to evaluate and choose a good multilingual knowledge base and documentation software?

Many of the existing support suites on the market include a documentation tool to create and run a knowledge base. Depending on the requirements of the Instruction Manager, this is a good and quick solution to use for the user manuals and knowledge base.

But handling multilingual knowledge and content is mostly only a small feature, if at all. For example, Zendesk and Freshdesk allow the creating of multi-language content, but the handling might not be perfectly suited to manage localized content.

If you have the possibility to choose a specific knowledge base or documentation product, the following criteria could be helpful:

• Does the product support all my required languages? E.g. languages with a specific alphabet or right-to-left text might be problematic.
• How are the content elements structured? Is there one content tree for all languages, or separate?
• How easy is it to translate the content? Can this be done directly by a (external) translator, or must this be done outside of the tool?
• If the main language content is edited, are the translated versions of the page marked as “fuzzy” that should be reviewed?
• How is the integration with other elements of the help center – e.g. to the release notes or the multilingual service portal?
• Can you assign the members of the multilingual support team to specific languages – e.g. define which employees can be used to translate, validate or support the requested languages?
• How does the public portal handle the localized articles? Is the url structure seo-friendly and all elements of the portal and content fully translated? If possible, the url should contain the keywords of the translated language, not for the main language of the portal.

Using automated translation with AI as gtranslate

Current, state of the art machine translation products as e.g. Google translate or other providers as Microsoft Azure have been improved a lot in the last years.

Computer-generated translation of articles and knowledge base entries is quite good and can be helpful e.g. for a initial translation draft that is reworked and finished by a human.

Since the development of artificial intelligence of such systems is getting better every day, it might only take a few years to fully replace the guilds of human translators.

Some of the limitations are handling of specific domain knowledge that might not always be detected and used by translation applications.

Need help to setup or run your multilingual knowledge base portal?

If you have a question or need some more information, reach out to us. Subscribe to our newsletter to get more information and content about multilingual documentation.