From API to XLIFF: Localization Terms You Need to Know

by | Oct 3, 2018

So you’ve decided to expand your brand globally! Among other tasks, this decision will require the translation of a number of different texts so your new audience will be able to understand what you’re selling.

But as you start your quest into the translation industry, you might encounter many words and phrases that you’ve never seen before.

Being met with acronyms and terms you’re unfamiliar with can be intimidating, so we’ve compiled a list of words you should know before you start the translation process. Read through from top to bottom or just come back to this page when you have questions —  it’s here for you!

Got a question about a term that’s not on here? Let us know at social@wordbee.com and we’ll add it to the list!

  • 110% match —  This is when a segment and the segment after it matches, and the context matches too.
  • API —  A “connector” of sorts on the backend of a system that tells a program to launch another action.
  • Application programming interface —  See API.
  • Backtranslation —  Another translation workflow. A text is translated from the source language to a target language. Then, the translated text is translated back to the source language. In doing so, workers in highly technical fields can more easily spot errors.
  • Bilingual evaluation understudy —  See BLEU.
  • BLEU —  Abbreviation for bilingual evaluation understudy. An algorithm to help determine the quality of a machine translation, providing an objective way to compare MT output.
  • Business analytics —  Measurements that show how translators and translation managers are performing.
  • CAT —  See computer-aided translation.
  • Character encoding —  A system of codes for each character of a language, including punctuation and symbols. Common encodings are ASCII and Unicode.
  • Computer-aided translation —  Abbreviated to CAT. A translation process in which a human translator completes a translation task with the help of a computer program (like Wordbee!). This may involve using translation memory, terminology managers, project management software, and other techniques. Also called computer-assisted translation.
  • Controlled English —  A set of rules for English that are used when writing something that will undergo localization and that make using translation memories easier. For example, when writing in controlled English you might avoid past or future tense, homonyms, or contractions.
  • Crowdsourcing —  A translation process in which a translation job is assigned to multiple translators who work on it at the same time.
  • End customer —  The person, group, or company that is buying the translation.
  • Exact match —  A segment in a text that matches a segment in translation memories, but not in the same context.
  • External translator —  A translator who works remotely, which benefits companies that don’t have a lot to translate or need translation services infrequently. This translator may or may not be a freelancer.
  • Freelance translator —  People who work for themselves for one or more companies at a time, and take tasks at-will. They might be referred to simply as “freelancers.”
  • Functional testing —  When someone tests an application after it’s been translated.
  • Fuzzy match —  A segment in a text that doesn’t exactly match up with translation memories.
  • Gig —  A project a freelancer might pick up.
  • Globalization —  The broadest decisions that can be made to something to bring it to new markets. It could include, for example, designing cars to meet a country’s strict emissions rules or creating offices in new countries to undergo local globalization processes. It’s often abbreviated “g11n.”
  • Group assignment —  A translation process in which a job is presented to a group of translators, and is assigned to the first one who takes it.
  • In-house translator —  A translator who works in your company’s office, which benefits companies that consistently have a lot of text that needs translated.
  • Internationalization —  When a software is adapted so it can be more easily put to use in new markets, which may involve changing the encoding as well as other details It’s often abbreviated “i18n.”
  • Localization quality assurance —  Abbreviated to LQA. Checks (usually made by an automated system) to find errors, such as if tags for formatting have been included or if the terminology is consistent. LQA can be done across projects, documents, and customers.
  • Language service provider —  Abbreviated to LSP. Businesses that take a translation job, give that job to a translator, and send the finished product to the end customer. The LSP guarantees the deadline and quality, and may mark up costs by 40%.
  • Localization —  When a text is adapted for a target market and may involve changing aspects of a marketing campaign or website to better suit new markets. It’s often abbreviated “l10n.”
  • LQA —  See localization quality assurance.
  • LSP —  See language service provider.
  • Machine translation —  A rules-based computer translation system. It tries to deduce rules for why words are translated one way or another, then applies those rules to the translation.
  • MT —  See machine translation.
  • Neural machine translation —  A form of machine learning in which an AI is applied to machine translation, looking for patterns and using them to create a translation. It’s similar to statistical machine translation.
  • Natural language processing —  Abbreviated to NLP. The ability of a machine to “understand” what something means. It’s used in search engines, as well as in text-to-speech and speech-to-text applications. NLP is applied in machine translation.
  • NLP —  See natural language processing.
  • Open API —  APIs that allow others to access the backend of a system from another system, like putting a different interface on a preexisting system.
  • Proofreading —  The process of reviewing a translated text, including the layout.
  • Pseudotranslation —  A technique that involves filling to-be-translated files with gibberish, then simulating text expansion. A pseudotranslation helps predict how web buttons and menus will work if the text of a document will end up longer or shorter after it’s translated. It also makes parallel work possible.
  • SDL —  One of the biggest companies that provides translation services, including a software. Its files aren’t always compatible with other files.
  • Search-engine optimized localization — See SEO localization.
  • Segment —  One sentence or one part of a sentence that gets stored for use in translation memories.
  • SEO localization —  The localization of specific keywords on a website, in addition to other techniques, to place a webpage higher among search results in a new target market.
  • Source language —  The language you’re translating from.
  • SRX file —  A file type that defines the rules of segmentation: what can become a segment, and what cannot become a segment.
  • Statistical machine translation —  A computer translation system that takes a huge set of data, then compares texts to find a translation that fits. It doesn’t understand grammar.
  • Target language —  The language you’re translating to.
  • TAUS —  An organization whose name stands for ‘Translation Automation User Society.’ TAUS comes up with standards for machine translation and language quality assurance.
  • TEP —  Abbreviation for ‘translating editing proofreading.’ A process used by many LSPs that provides quality assurance. In this process, a translator translates the text, an editor checks it against the source text, then a proofreader checks the language and layout.
  • Terminology management —  A company that creates a style guide system, which will show terminology terms while a translator translates.
  • Text analytics —  A subset of natural language processing in which a computer takes a piece of text and runs analytics on it. These analytics allow it to detect certain pieces of information like the topic, any proper nouns, and important terms, and even create a summary.
  • Text contraction —  When a text gets shorter after it’s translated into a new language.
  • Text expansion —  When a text gets longer after it’s translated into a new language.
  • TMS —  See translation management system.
  • Translation agency —  See language service provider.
  • Translation editing proofreading —  See TEP.
  • Translation management system —  This is a system used to manage a project and its supplies, finances, and workflows.
  • Translation memories —  Previously translated segments that will get reused by a translation program.
  • Transliteration —  When you change characters of a language from one kind to another, as would happen from languages like Japanese and Russian.
  • Unicode —  This is an international character coding system.
  • Workflows —  This is the process that a translation project will undergo.
  • XLIFF —  The standard format for the exchange of translation files. In theory, translators should be able to edit on it then interchange it between other file formats.

Is this interesting?

Subscribe to get interesting localization podcasts, discussion panels, and articles every month.

Wordbee all-round features

Reaching global markets means you need to get your translation management together. 

Want cool localization techniques straight to your email?

Keep up on the latest in localization management techniques with Wordbee.

You have Successfully Subscribed!