Spanish Translation Blog: Spanish Translation US » Blog Archive » Crowdsourcing Translation

Crowdsourcing Translation

, also known as massive outsourcing or voluntary subcontracting, is outsourcing a task, that was traditionally done by an employee or a contractor, to a large group of people or a community through an open call. As of late, the Internet has used it to introduce and manage new projects.

This is how it works: there is an open call requesting the help of volunteers. The users, known as the mass, usually form online communities to help out a company and review everyone’s suggestions to an issue. Once the best solution is agreed upon, the company incorporates it. This process is known as crowdsourcing.

Crowdsourcing translation is a big trend now, since it can help companies boost their brands and position within the market. However, both the term and the underlying business model have generated criticism and controversy.

As noted in an article in Softpedia, there are companies that believe that crowdsourcing breaks all the boundaries associated with the traditional , high costs, constrained resources and the inherent errors caused by discrepancies between language skills and knowledge on the subject.

In crowdsourcing, a collaborative translation platform is used that serves as a special translation tool. It is a global web based platform that facilitates the translation process and allows companies to reach a global audience, as well as encourages participation and increases brand loyalty within these communities.

The content introduced in the platform is offered to the community as a translation workbench and shows the source text for the users to translate. Translators have several resources available, such as translation engines and a statistical engine that provides a machine translation and gives an overview of the content, although it is not of high quality.

The translation memory is another resource used. Every time a sentence is translated, it is stored in both languages. When this system is used or data is loaded, and the original content reappears, the platform recognizes the segments that were already translated. The system allows users to reuse translations made by other human translators with the translation memory.

So far, companies decide what content is available in a foreign language. Indeed, this is what crowdsourcing is trying to avoid: the community can select the content they want to translate, and consumers can select the content they want translated.

Is this better or worse than the translation?
Will it work better or more result in faulty and inaccurate translations?
Or are human translators the most reliable quality wise?