Research

Jennifer's current field of research is a rather unlikely subject for a professional translator: controlled languages and machine translation. 

Controlled natural languages (CNL) restrict the grammar and vocabulary of an existing language, such as English, in order to reduce or eliminate ambiguity and complexity. They are used to facilitate understanding by human readers, or to enable documents to be translated by machine (computer-based) translation (MT) tools.

Research has shown that CNLs substantially improve the quality of translations delivered by MT tools. However, they have hitherto been developed almost exclusively by major international corporations to facilitate the machine translation of their technical documentation, and little or nothing has been done to adapt them to public domain MT tools. 

Google Translate is without doubt the most widely used MT system today, and arguably the best in certain technical fields. Would it not be possible, therefore, to develop a CNL for use with Google Translate, and thus harness the real potential of this highly sophisticated tool?

Intrigued by this thought, Jennifer developed an experimental CNL in Spanish for translating the abstracts of medical articles into English using Google Translate for her Master's dissertation. This is the first time a CNL has been developed for use with a statiscs-based machine translation tools, such as Google Translate, and also the first MT-oriented CNL developed in Spanish. The results have been excellent, and Jennifer is currently engaged in perfecting and widening the scope of the CNL and exploring the potential of this new approach in other fields.

 

 

Call

T: 34 609 817 818   T: 46 707 589 289
 

  • facebook
  • Twitter Clean
  • w-googleplus

Follow me

 

Jennifer Gelfgren