BOCA RATON, FLA. -- Speech technology will be the "killer app" when it comes to the future adoption of mobile applications, according to a leading technologist's presentation here this week at Conversations, the user conference for speech technology vendor Nuance Communications.
Citing statistics of mobile-Web adoption rates, Vlad Sejnoha, chief scientist at Nuance, presented a case -- or, more accurately, a challenge -- for the rapid, cutting-edge development of speech-enabled applications for mobile services. Focusing specifically on local search and Web applications, Sejnoha laid out a roadmap of what needs to happen for his company's mobile vision to be achieved. His presentation, "Speech is the Catalyst," also gave equal weight to the ever-changing needs of end users. "As we plan the next generation, we are increasingly aware that we are catering to users whose expectations are changing," he told the audience.
But, he stated, no speech-powered mobile applications can come to fruition and achieve wide adoption without innovation. Though mobile-phone penetration has reached 80 percent -- and 80 percent of those devices are Web-enabled -- Sejnoha said the mobile Web is "lagging" compared to that of the general Web. "A lot of what's available from mobile is not usable," he told the crowd. Because the mobile Web is still fragmented -- users must complete a variety of steps to reach their destination -- many consumers prefer not to use it. The consumer, Senjoha said, demands faster results than those being delivered by today's technologies.
Sejnoha presented the audience with his version of how speech applications could boost mobile-Web user-adoption rates: Speech, he said, is the answer for a consumer who wants fast, easy Internet access. In a video produced by Nuance, two people walk through a city, trying to pick a restaurant for lunch. One user takes out his iPhone, requests information and reviews of a restaurant from Yelp.com using voice commands, and reads through the information. Afterward, he dictates his own review of the restaurant through his iPhone, with Nuance technology providing the transcription. Though not possible today, Sejnoha says this future will "become prevalent not too many years from now."
Sejnoha also urged the audience to be honest about certain mobile Web functions. Consumers, he said, may outgrow certain applications, but any program will require continuous innovation to remain both popular and relevant. Furthermore, future applications looking to gain an edge over competitors will have to provide what consumers will be clamoring for: more accurate automatic speech recognition (ASR), as well as the ability to search for a topic with additional spoken parameters. This, Sejnoha said, is another avenue for speech penetration into the mobile arena.
"Mobile search has the promise of becoming the same gateway to the Web as [browser-based search has been for] desktops, but [mobile search] is hindered by the difficulty of text entry," Sejnoha stated. "Given freedom, people want to describe their problems or searches more precisely than [through] a couple of key words." The ability to perform more text-heavy operations is where speech recognition and transcription will play a vital role.
Of course, without quick responses mobile Web adoption will not succeed, Sejnoha said. The remedy, he told the crowd, is natural language, and the ability to exract meaning from utterances. "You might be interacting with a number of services, but might not know what they are," he said. "Natural language lets you implicitly request the service."
And though speech-enabled applications in mobile phones have yet to be fully realized, Sejnoha said he remains confident in speech's ability to enter the market, and to eventually create a significant customer demand.
"Maybe our aspirations are more high-reaching," he said. "Just like the Web really benefited from general search, we think that the mobile Web is poised for transformation technology, and we think speech has the potential for being that catalyst."