It is available in the UAE on Android 2.2+ and iPhone
Google, the world's most popular search engine, has launched voice search in Arabic to help people convert data and information into knowledge, said a press statement.
With users increasingly accessing the web through their mobile phones, Voice Search facilitates the process by allowing users to search by simply using their voices. Voice Search uses Google's speech recognition technology to turn the spoken word into text and run a search on Google just as if the query was typed by hand.
Voice search is available in the UAE on Android 2.2+ and iPhone.
"Google is leading the way in mobile and speech recognition innovation," said Najeeb Jarrar, Mena Product Marketing Manager. "Our universal search technology makes information accessible to users all over the world and Voice Search in Arabic is an important step in allowing users greater access to these innovations across the Mena region. This new service is just one more way that Google is showcasing the remarkable possibilities of the Internet."
Voice Search is powerful and fast, because all the computing takes place in mere seconds through Google's datacenters. The three-step Voice Search process starts when Google streams sound files to their datacenters in real-time. They then convert the audio information into phonemes, into words, into phrases. Finally, Google then compares phrases against the billions of daily queries to assign probability scores to all possible transcriptions and delivers results.
Continuously committed to providing Google's cutting-edge applications worldwide, Google's engineering team worked to combine local expertise with their global speech recognition technology to bring Voice Search to the UAE.
The speech recognition technology that powers Voice Search was developed entirely in-house. Google experts taught computers to understand the sounds and words that make up spoken language to form the foundation for speech recognition. To build Voice Search Arabic, Google worked directly with native speakers in the UAE, Egypt, Lebanon, Saudi Arabia, and Jordan to collect speech samples on the ground.
Google recruited local staff who, in turn, recorded local native speakers. Volunteers were asked to read popular queries in their native tongue, in a variety of acoustic conditions such as in restaurants, out on busy streets, and inside cars. The team spent weeks collecting spoken words to create the specific models which power the service. The results were then sent to Google speech scientists in the US worked to refine the speech models.
© Emirates 24|7 2011
