If you’re new to the translation and localization industry, or are just shopping for a solution, you’ll be confronted with a lot of jargon. This glossary explains this jargon in terms that relate to ordinary businesses. Of course, if you have any questions, we’re here to help.
API (Application Programming Interface)
An API is a interface that other computer programs communicate with to make requests. In the context of translation, an API enables an application, such as an online store web server, to call over to a translation service to request a translation for something, and then automatically receive a response when it is done. Essentially what a web API does is define a standard way for other programs to submit requests and receive responses.
CAT Tool (Computer Assisted Translation) Tool
Computer assisted translation is the application of automated systems, such as machine translation engines and translation memory, to assist translators in working more efficiently, and in producing consistent output quality. Most translation agencies, and most freelance translators, use some type of CAT tool, such as desktop programs like Trados that combine editing tools, connections to machine translation and translation memory. Your translation service provider will generally provide these tools to translators working on your projects.
CMS (Content Management System)
A content management system enables non-technical people to write, edit and publish material, typically without needing to know how to code web pages in HTML. A good CMS enables many people to work in parallel, and to work in different roles (e.g. author, reviewer, editor, translator, etc). Modern content management systems also support multilingual content, so a website can be published and maintained in many languages. If your website is more than a few pages, you need a CMS. If you are publishing your site on a hosted environment (someone else is taking care of hosting the site for you), you’re probably already using one.
A translation corpus is a collection of source texts and their translations. This is typically used to build translation memories (a collection of previously created translation that help translators deal with repeating phrases or terminology), and to train machine translation systems such as Google Translate. This is not generally something you need to think about unless you are managing a very large volume of texts and translations.
A glossary or phrase glossary is a dictionary of terms and their translations. Customers frequently use glossaries to insure that certain words, phrases and proper names are always translated consistently. This is especially important when multiple translators are working on a project because each may translate the same phrase in slightly different ways. A glossary prevents that. Most translation workbenches include glossary support. Glossaries can also be used to reduce translation cost, so repeating phrases are only translated once, although this has less of an impact than you might think, a few percentage points in cost reduction at most.
HTML (Hypertext Markup Language)
HTML is the markup language used to create web pages. It is so simple that it can be edited by hand, although most people create web content in “WYSIWG” (what you see is what you get) text editors that hide this from view. Most translation services can process HTML, and know how to translate human readable text while leaving the HTML code alone.
Hybrid translation is a process that combines machine translation, crowd or user generated translations, and professional translation. This technique enables you to optimize for speed, quality and cost. For example, machine translation is instantaneous, and can be used to generate temporary placeholder translates for a news article while the professional translations are being done, and then replaced once the pro translations come back. This is especially important in high volume applications where the cost of professionally translating every text is prohibitive.
LSP (Language Service Provider)
LSPs or language service providers are also known as translation agencies. They recruit, vet and manage a pool of translators, anywhere from a handful of people to tens of thousands of freelancers. Their primary role is to recruit and manage translators. If you need support for languages your staff does not speak, or have a large volume of material to translate, an LSP can help you greatly.
TDN (Translation Delivery Network)
A translation delivery network is a service that sits between end users and your website. The TDN “redraws” web pages as they are served, and inserts translations to replace the original language text, all while preserving the original look and feel of the page. This technique eliminates the need to install software on the original web server, and can be rolled out in a day. Examples of companies that offer this service are Smartling and Dakwak.
TM (Translation Memory)
A translation memory is a database of previously created translations. This is used to assist translators in presenting a consistent style and voice to the reader, by giving them examples of similar translations. There are two types of translation memory: exact match and fuzzy matched. In an exact match system, only identical source texts and their translations are retrieved (this is used to prevent re-translation of texts that have already been translated before). In a fuzzy match system, similar source texts and their translations are displayed. This enables translators to see translations for source texts that were similar but not identical, which is useful in showing them examples of how others before them did their work.
TMS (Translation Management System)
A translation management system automates much of the project management associated with translation projects. A TMS, for example, shows translators the queue of pending tasks within a project, and enables many translators to work in parallel. The system also enforces whatever review and QA procedures the owner has specified, for example to require independent reviewers to score each translation before it is released to the customer. The primary value of a TMS is to streamline work, and to minimize the amount of non-translation related overhead associated with a project (in a traditional agency, project management accounts for up to a third of overall project cost).
Unicode is a universal character encoding standard. It is similar to the ASCII character set, where each letter is represented by a numeric code, except that it has addresses, or code points, to represent every character from every alphabet in use today. This standard is supported by all web browsers and most applications. Without it, developing useful multilingual applications and services would be almost impossible.
XML (eXtensible Markup Language)
XML is a structured document format that enables users to easily share structured data, such as a set of records from a database. XML is widely used as a way to share translation records and related information between translation services.