What do WordPress, Linux, and Firefox all have in common? All of these successful projects are the result of crowdsourced contributions. As the Internet continues to connect us, collaborating on projects has become easier than ever. Even in the localization industry, crowdsourced translation solutions are helping to make translation services available to everyone and the advent of new technologies has brought about a few different methods of collaborative translation projects. Let’s take a look at three of the most popular models of crowdsourced translations.
Machine Learning
Machine learning is one of the hottest trends in the translation world right now. As artificial intelligence improves, different machine translation (MT) engines are taking advantage of modern technological advances to become more dynamic and responsive. While these systems are promising amazing results—predicted to soon be on par with some of the best human translators – they require a significant amount of effort to set up. In addition to spending years coding and developing complicated algorithms, the human effort required to teach these programs is also significant. Thousands of people must review different strings of text and their translations to determine whether or not translations are accurate. The quality of the human feedback is also very important, as poor-quality input can end up leading to poor-quality output from the MT engine later on. While machine learning may be the future of translation, it is probably too costly and labor intensive for the average company looking for an easy solution to its translation needs.
Community Translation
In contrast, community translation provides a cheap and easy way for organizations to help translate their content. In communities that have strong investment from members, it can be easy to find plenty of linguists who are willing to donate their services. Online gaming communities, such as Second Life and Steam, have great translation communities that help them translate into dozens of different languages. While translation communities are often free, they often sacrifice quality. This is fine for “low-stakes” translations such as games or online communities, but may not produce the level of quality needed for technical or industry-specific content. If you are looking for a service to help with these “high-stakes” translations, you might be better off exploring other options.
Crowdsourced Translation
Crowdsourced translation involves large numbers of people contributing to projects; similar to community translation, but with better management and quality control. With crowdsourced translation, translators are managed and appraised by a language service provider (LSP). The LSP manages client feedback and ratings, keeping track of which translators receive the best feedback, and ranking them in order to ensure quality. LSPs can also keep track of which translators are qualified to translate technical information, such as content related to aerospace or the life sciences industry. This helps ensure that your “high-stakes” translations are done right, every time. While the LSPs keep track of translators, they also provide easy-to-use interfaces, such as Stepes, which allow you to communicate directly with translators. This allows you to cut out the middle man, saving you time, effort, and money. Solutions like Stepes allow you to reap all of the benefits of big translation, without sacrificing any of the quality.
When looking for a solution to your translation needs there are a lot of different approaches you can take. Professional, crowdsourced translation services offer a balanced solution that ensures high-quality translations, while also offering an affordable price and convenient access to translators. Stepes, the world’s first chat-based translation app, is a great example of a crowdsourced translation solution that can help with your translation needs. Visit stepes.com today to find out more information!
Written by John Wawrose – Senior Technical Writer at CSOFT International
read more blogs by John Wawrose