Nuance has released a new voice search app called Dragon Go, for the iPhone and iPod Touch, now available on the Apple App Store. Its reach - the app can search some 180 mobile sites - illustrates one facet of the advances this technology is making, for U.S. users. Other examples can be found in further flung global markets.
Dragon Go searches such sites as AccuWeather, Bing, ESPN, Facebook, Fandango, iTunes, Last.fm, LiveNation, Milo.com, OpenTable, Pandora internet radio, Rotten Tomatoes, Twitter, Wikipedia, Yelp, YouTube and Yahoo. The app is able to complete this process - spoken request and then the return of results - via a combination of natural language understanding, analytical tools, and artificial intelligence.
Typical Use
A typical user might say "The best Mexican food near me," and be taken to the Yelp tab for local reviews. Or he may say "Reservations for two at Capital Grille in Boston." The app takes the user to OpenTable to book a table for dinner. It also combines the search with other functionality. If someone asks for the number of a local store, the user can then tap and call it when it is pulled up. A request to "Lady Gaga on Pandora" let the user hear her personalized Lady Gaga station.
On the Desktop
Google, not surprisingly, is driving much of the innovation in this space, most lately in June when it ported the technology to the desktop. For the moment, though, it appears voice search development and use is mostly centered around the mobile use case. Dragon Go, for instance, followed a similar app released earlier this year by SoundHound. Called "Hound“, it is a voice-search application for music lovers. Users find a song, artist, or band by saying it. Up comes results that show artist page, complete with previews of top songs, lyrics, videos and sharing features. It is also available through the App Store, through Amazon's Appstore and Google's Android Market.
Going Global
Perhaps more interesting are Google’s efforts to offer Voice Search in other markets and languages. Earlier this year it launched Voice Search in Indonesian, Malaysian, and Latin American Spanish, making the application available in over two dozen languages and accents since the first launch in November 2008. Google's been launching the product in foreign markets via the help of local users in the region, a different approach, it says in its blog.
"The traditional method of acquiring voice samples is to license the data from companies who specialize in the distribution of speech and text databases." Google, instead, opted to tap local communities giving volunteers phones to get voice samples from their friends and family. "Everyone was required to sign a consent form and all voice samples were anonymized. When possible, they also helped to test early versions of Voice Search as the product got closer to launch."
It took thousands of hours of raw data to capture regional accents and idiomatic speech in all sorts of recording environments to mimic daily life use cases, it said. "For instance, when developing Voice Search for Latin American Spanish, we paid particular attention to Mexican and Argentinean Spanish. These two accents are more different from one another than any other pair of widely-used accents in all of South and Central America."
