LINGUISTS and computer scientists at a Scottish university have launched an app to to teach computers to “speak” Scots Gaelic.
The device helps computers understand Gaelic text, can be used in a range of functions such as voice recognition and online translation, as well as grammar and spell checks, and has applications that will assist people to understand and learn the notoriously-tricky tongue.
Researchers at the University of Edinburgh used a database of more than 100,000 Gaelic words to develop an app -- a so-called a part-of-speech, or POS tagger, which allows computers to pair a word with its grammatical label, such as a noun, verb or adjective.
POS taggers have already been developed for a number of other languages, but this is the first such language technology to be tested for Scottish Gaelic.
The online tool helps computers to correctly categorise and order a language - a key process in digital applications, such as computer generated speech.
The product has been developed by researchers at the University of Edinburgh, with support from the Universities of Aberdeen and Glasgow, Edinburgh Napier University and Sabhal Mor Ostaig in Skye, part of the University of the Highlands and Islands.
The project, which is funded by the Carnegie Trust and Bord na Gaidhlig (The Scots Gaelic Board) was launched at a Gaelic research conference in Skye.
Dr William Lamb, of the University of Edinburgh’s School of Literatures, Languages and Cultures, said: “Corpus and speech technology in Scottish Gaelic remains in an embryonic stage, in contrast to the leaps made for other languages in this area in recent decades. This project provides the means to redress the situation and brings the language forward - with a vital technological and intellectual push - into the 21st century.”