Sherehe za Queen Elizabeth's Diamond Jubilee zaambatana na Uzinduzi wa application ambayo itamwezesha mtu kuandika Msg na ku Tweet kama Queen

(Reuters) - In celebration of Queen Elizabeth's Diamond Jubilee, a mobile app is adding a royal touch to emails, texts and tweets by suggesting words that the British monarch would be most likely to use.

SwiftKey is an Android app that uses artificial intelligence to correct and predict words as they are typed. To mark the Queen's 60-year reign, the company released the Queen's English, a new language module for the app.

"One of the core strengths of the technology is that we are able to take any kind of text and build what we call a language module," explained Dr. Ben Medlock, co-founder of British company SwiftKey.

"It solves a probabilistic problem which is ‘what is the person most likely to say next'?"

The company created the module using its underlying language technology, which processed the transcripts of Queen Elizabeth speeches since her coronation in 1952.

It revealed several trends, including the Queen's avoidance of colloquial contractions such as ‘it's' in favor of the more formal ‘it is', and her tendency to maintain a positive tone in communications.

"The Queen's language reveals that she has a generally optimistic frame of mind and so the words ‘confident', ‘delighted', ‘glad' and ‘please' were uttered 125 times more than her famous ‘annus horribilis,' which was the phrase that she used in the 1992 Windsor Castle fire," said Medlock, who has a PhD in natural language processing from the University of Cambridge.

The app was conceived two years ago on the premise that the problem with typing on smartphones lies in language, rather than in keyboards.

"Smartphones were exploding but people were struggling to do what is thought to be the most important thing you can do on a phone, which is to get your thoughts down into it," he said.

"We realized that the software of the future wasn't just going to sit there as a dumb keyboard based on key strokes. It was going to actively model the way people use language."