Advertisement

Modern language experts at the University of Leeds are developing computer programs which could alter how languages are taught and used around the world.

Currently, many computer-based systems for teaching and translating languages are out of date or not user-friendly. While there is a wealth of information about different languages and vocabularies available on the internet, much is badly-organised and difficult to navigate.

Dr Serge Sharoff of the Centre for Translation Studies is working on three research projects - Kelly, TTC and Accurat - to bring things up to date. He has been awarded close to £700,000 from the European Commission in recent months to fund the work.

In the Kelly project, a set of word learning cards for some of the most frequent words in Swedish, Norwegian, Greek, Polish, Arabic, English, Chinese, Russian and Italian will be developed, along with a scientific basis for measuring learners' abilities in these languages.

According to Sharoff, some language teachers are unhappy with the word learning cards currently available because of inconsistencies in grading. The Kelly project is designed to produce word lists from large collections of data and adapt them into a new, learner-centred framework.

Dr Sharoff said: "Flashcards are useful for language learners but need to be graded in terms of difficulty. This grading exists but hasn't been done rigorously until now. Words will be classified individually and by 15 subject categories - such as food and drink, nature and animals - into a visual scheme. This will allow learners and teachers to deal with words systematically, by setting clear aims, planning studies and controlling progress."

TTC (Terminology Extraction, Translation Tools and Comparable Corpora) aims to provide new tools for translators by producing up-to-date term lists from original texts in French, German, Spanish, Chinese and Russian for rapidly developing areas, such as wind energy or mobile phones. This will be done by taking a variety of different methods utilising a large amount of original texts in these languages. Such term banks can be used for translation, including, including machine translation - using computer software to translate speech or text from one language to another - and computer-assisted translation tools - translators using computers to assist in their work - with multilingual content management tools.

The third project, Accurat, will research new methods and techniques to overcome one of the main problems of machine translation - the lack of linguistic resources for languages which are relatively under-represented on computers, such as Latvian, Estonian, Romanian, Greek and Croatian.

Dr Sharoff said: "The main goal is to find and evaluate new ways to exploit the properties of independently written texts in some languages in order to compensate for the shortage of translation resources in others between them. Ultimately, this is to improve translation tools for languages that are currently poorly-served."

For more information contact:

Guy Dixon, University of Leeds media relations - 0113 343 8299 or g.dixon@leeds.ac.uk

Notes to editors

The 2008 Research Assessment Exercise showed the University of Leeds to be the UK's eighth biggest research powerhouse. The University is one of the largest higher education institutions in the UK and a member of the Russell Group of research-intensive universities. The University's vision is to secure a place among the world's top 50 by 2015. www.leeds.ac.uk

In the most recent national review of research quality (RAE2008) CTS@Leeds researchers were assessed together with colleagues in Computer Science and Informatics. The review concluded that 80% of the research outputs were 'world leading' or 'internationally excellent', while the remaining 20% were 'internationally recognised' - these being the top three grades on a five-grade scale.

SOURCE

Advertisement
Advertisement