COPYRIGHT 2000 Andrew Seybold's Outlook, Inc.
With the Wireless Applications Protocol (WAP) and all of its derivatives doing poorly in the Americas and in Europe, many folks are wondering whether this business called the Wireless Internet will be such a big deal after all. I-mode has more than ten million subscribers, but as I have pointed out before, there are many differences between Japan and the U.S.--the most significant of which is that 85% of all i-mode users have never seen the Internet on anything other than their phone. It will be interesting to watch what happens to i-mode growth as NTT (the parent) rolls out high-speed data access over its wired network and customers in Japan are exposed to the Internet as we know it.
In the meantime, it is becoming clearer every day that voice access to the Internet and to corporate information will be an important factor in the wireless industry, and many companies are rushing into this space. Even Phone.com, the company that pioneered WAP in all of its glory, has purchased a voice access company (@Motion Inc.). America Online recently acquired Quack.com for a reported $200 million, and more than a few startups have launched voice services. Companies such as BeVocal.com, ShopTalk.com, and SwiftTouch are ready and willing to provide voice access to information services. Some of these companies, including Indicast, a San Diego-based startup, combine voice recognition with Internet audio feeds. If you ask for news about IBM, they retrieve a feed that provides real audio for you to listen to.
Everypath provides wireless access to both Internet and corporate sites with support for WAP and other data formats as well as voice. Motorola is spearheading a consortium of companies that are interested in voice access and is driving a voice-to-internet standard call "VoxML" (Voice XML). Several hundred companies are now involved in this project. Companies such as Oracle with its Portal to Go products are also enabling voice access to information.
The Race Is On!
Voice will be the next big thing, but a lot of work remains to be done before voice will be a truly ubiquitous access method (see Barney Dewey's Threshold of Usability model, vol. 1 g, no. 1, August 2000). Voice recognition technology is not yet "user-proof," though it is getting better, and it can sometimes be frustrating to interact with--especially when background noise is picked up by the phone's mouthpiece. Systems that work well on a desktop fall on their faces when the user is standing at the gate at an airport and the overhead speaker starts blaring a boarding announcement. Using voice access from a car with the windows down, the AM/FM radio on, or even, in some cases, the fan from the air conditioner or heater turned up high can turn the user experience into a real nightmare.
Early adopters will put up with a lot in order to be using the latest and greatest gadgets, but in order to cross the chasm and get the technology into the hands of the general population, we still have a lot of work to do. Before we delve into activities surrounding voice recognition technologies and services, let's define the different levels of speech recognition as we know it today:
* Command and control
* User-specific speech recognition
* User-independent speech recognition
Read the full article for free courtesy of your local library.