As a translation or localization newbie you might not be familiar with all the abbreviations and terms used in the business. That’s why we’ve set up this #l10n glossary to give you a quick overview of the most commonly used terms and help you getting started.
Continuous Translation is the opponent to traditional translation processes. It integrates translation smoothly in agile development process. Typically Continuous Translation features
- a high drop frequency of projects
- agile and fast translation cycles
- a small translation volume
- a quick turnaround.
Exact matches are suggestions from Translation Memory which suit a given text completely. These matches correspond to the actual text to 100%. Hence, suggestions don’t need to be adapted and can be used as final translations.
Fuzzy matches are rough matches suggested from Translation Memory. They suit a given text between 55% and 99%, thus, they need some correction work in order to create final translations. Fuzzy matches are a great starting point to work out high-quality translations.
Internationalization describes the process of extracting all texts of a software or product and placing them into so-called resource files. It is the basis for fast localization and translation of software.
The term is often used in localization and consists of a set of parameters that defines a user’s language and country or region. In translation and localization projects a locale typically consists of a language identifier and a region identifier.
Localization describes the process of adapting a product, application, software or document to a local language, culture and other specific requirements of a country or region. Besides translation localization also includes the adaption of time, date and numeric formats, currency, payment methods, images, colors and more.
Machine Translation, also called automatic or instant translation, is the text translation by a computer without human involvement. There are many concerns about translation quality because it bases on computer algorithms and translate text without human assistance.
Minimum viable Localization
MLV is the equivalent to the minimum viable product approach in development. It’s typically used in mobile app localization and refers to a strategy for testing different markets and their potential for an application. You localize your app store description and metadata for various international markets without translating the application itself. Minimum viable localization is a great way to test certain markets at low costs.
In terms of translation quality Machine Translation cannot keep up with human translators so far. Post-editing describes a way of using Machine Translation and human translators to accelerate translation. Texts are machine translated and then edited by professional translators to work out high-quality translations in minimum time.
Language files you use for an application’s or document’s source texts and to store translations are resource files. Typically resource file names include the language abbreviation or locale to classify the file’s language.
RTL represents the abbreviation for Right to Left that refers to language that you write and read right to left, like Arabic or Hebrew. Translating your application or document from an LTR language (like English) into a RTL language comes along with considerable changes in layout. Hence, RTL support is important to consider in internationalization.
A Term Base, also referred to as glossary, is the primary tool for terminology management in translation projects. It’s a compilation of company or industry specific terms and typically consists of
- Terms (source language)
- Approved translations (target languages)
- A description
- Terms that are not to be translated but kept in the source language.
A Translation Memory is a database that learns and grows along with every translation you save. It suggests translations for given texts in real time based on similarity with past translations (fuzzy or exact matches).
Unicode is an international computing industry standard used for text representation and handling in most writing systems. The most commonly used encodings are UTF-8 and UTF-16.
We missed any localization terms a global company definitely needs to know? We’re happy if you share them with us in the comments!