Crowdsourcing could help improve predictive texting

London, Feb 26 (ANI): Smartphones may soon be equipped with a much better predictive text system - thanks to the wonderful concept of crowdsourcing, researchers have suggested.

The software packages installed in present day mobiles usually struggle with texts and voice commands in case a user tries to use words or phrases that are not available in the phone's database.

To determine if the crowd could help, Keith Vertanen of Montana Tech in Butte and Per Ola Kristensson at the University of St Andrews, UK, paid 298 workers of the Amazon Mechanical Turk to envision phrases they might require if they suffered from motor neuron disease or cerebral palsy.

The plan was to try and improve a predictive system used in Augmented and Alternative Communication (AAC) devices, which help disabled people to communicate by painstakingly typing out words interpreted from their muscle twitches or blinks, the New Scientist reported.

The worker's responses generated almost 6000 constructive phrases. Next, by sifting through Twitter posts and other social media texts, the participants extracted sentences and phrases with related structures to the initial phrases, escalating the quantity to tens of millions of entries.

The result is a system that requires 11 per cent lesser keystrokes than a standard AAC device.

Kristensson asserted that it would be a huge improvement for anyone struggling over every word.

It should also work when applied to more standard text and speech recognition systems, he added.

The work appears in the Proceedings of the 2011 Conference on Empirical Method in Natural Language Processing. (ANI)