Date post: | 14-Aug-2015 |
Category: |
Documents |
Upload: | bui-danh-nhan |
View: | 19 times |
Download: | 0 times |
4Verizon Communications, AAA,and Others: Advances in SpeechRecognition Software Are Extendingthe Utility of Traditional Applications
The velvety voice of that nice young woman on theother end of the phone is really just digits on a disksomewhere at Verizon Communications Inc., but
"she"remembers that you spoke to her a few moments ear-lier,beforeyou were interrupted. "I apologize if I ask somequestionsyou already answered," the voice says. She soundsgenuinelycontrite.
But the virtual telephone-repairwornan is just gettingwarmedup. "I'll test your line from here," she intones. "OK,Igotthe line test started. It could take up to a minute. I'llalsocheckto see if anything's changed on the line since youlastcalled."While the test runs, she asks you for more infor-mationabout your telephone problem, and she seems to un-derstandyour every response.
Presentlyshe says, "The line test is finished now. Unfor-tunately,it couldn't determine if the problem is in Verizon'snetworkor with your equipment, so we need to dispatch atechnician.... Here we are-I've picked up all of our tech-nicians'current schedules. The earliest we can schedule it isonThursday,June 3, between 8 a.m. and 6 p.m. Can some-onegiveaccess to the premises at that time?" The call issooncompleted, and on June 3, so is the repair.
Computerized speech has come a long way in 20 years.AsVerizons system illustrates, the technology has becomesmarter,easier to use, and more integrated with other appli-cations.Such technical advances, plus product introductionsthatfacilitate the deployment of the technology by main-streamdevelopers, are enabling new uses for automatedspeechsystems.
Researchin automated speech recognition (ASR) goesbackto the 1930s, but serious commercialization of it didn'tbeginuntil 50 years later. In 1988, Dragon Systems Inc.demonstrateda I'C-based speech recognition system with anS,OOO-wordvocabulary. Users had to speak slowly andclearly.One. Word. At. A. Time.
The next big step came in 1990, when Dragon demon-trated a 5,000-word continuous-speech system for PCs anda large-vocabulary,speech-to-text system for general-purposedictation.Then, in 1997, Dragon and IBM both introducedcontinuousspeech recognition systems for general-purposeuse.
Meanwhile, corporations began rolling out interactivevoiceresponse (IVR) systems. The earlier ones-indeed,mostin use today-are menu-driven: "For your fund bal-mce,sar or press 'one.'" A few advanced systems are moreconversational:"What city are you departing from?" Despitethe steadyadvancements to bigger vocabularies, lower errorrates,andmore natural interfaces, however, speech productshaveremainedspecialized tools for niche markets such as PCDa\;gationby the disabled, medical dictation, and tightly
. . erviceinteractions,
and to spawn transactions. As a result, these speechsystems-previously the domain of call center and telephonymanagers-are increasingly becoming something for the ITshop to worry about, if not manage.
Verizon's speech application, for example, can trigger aline test, update customer accounts, schedule repairs, andcreate trouble tickets-processes that require interfaces withmany systems. "If you create something that's just a veneer,people get it very quickly," says Fari Ebrahimi, senior vicepresident for IT at Verizon. "But for customers to really getvalue, you need to do something with the back office."
Many of Verizon's back-office functions have been re-designed as Web services and are accessible by customersover the Web or by spoken request. The new system handlessome 50,000 repair calls per day and has boosted the per-centage of calls that are fully automated from 3 percent to 20percent, Ebrahimi says. He won't say how much the com-pany is saving in labor costs, but he says it's "millions andmillions. "
Verizon's National Operations Voice Portal is deployedacross three geographically dispersed data centers, and callsare routed from point to point using voice-over-If' technol-ogy. The system uses speech recognition products and userinterface designs from ScanSoft Inc. (which obtained much ofDragon's speech technology via acquisition). Telephonyservers at each data center are connected to back-office appli-cation servers running BEA Systems Inc.'s BEA WebLogicServer.
"The technology that used to be in those telephone silos,managed by the call center manager, is now becoming stan-dards-based and is being driven by the same applicationserver that serves the Web pages," says William Meisel, pres-ident of TMA Associates, a speech-technology consultingfirm in Tarzana, California. "Now the IT department cancreate the applications in an environment that's more famil-iar to them."
Organizations that have deployed speech technology saythat recent advancements in natural-language understandinghave made the systems more acceptable to callers. "WithIVR, it was 'Touch or say three,'" says Joe Alessi, vice presi-dent for marketing and IT at AAA Minnesotallowa. "Nowwe can say, 'I'd like to change my address.'''
The organization last year replaced a Touch- Tone-basedIVR member service system with a self-service system builton the Say Anything natural-language speech engine fromNuance Communications Inc. One objective was to reduceturnover in the call center by freeing agents from handlingmundane calls, such as requests for new membership cards.Another goal was to address the problem of callers bailingout of the IVR system because they found the menus confus-ing, Alessi says.
tal annual savings of $200,000, according to AlessI.T. Rowe Price Group Inc. in Baltimore also upgraded
its menu-driven IVR system to a free-form speech systembased on IBM's WebSphere Voice Response and VoiceServer with natural-language understanding capabilities.The investment company reports big savings in telephonecharges because automated calls can be completed faster.
CASE STUDY QUESTIONS
1. What are the business benefits and limitations ofIVRat Verizon and others? How could their use of IVR beimproved?
2. What types of business situations would benefit mostfrom IVR technology? Which ones would benefit least?
3. Given the advancements in voice recognition softwareover the last 20 years, what types of new applicationsfor IVR do you see in the next 20 years? Give examples.
nee. ey cou a ta e tree to four minutes, because yhave to go through five or six menu legs. Now the sametransaction takes 30 seconds because you can speak it all inone sentence."
Source: Adapted from Gary H. Anthes, "Speak Easy," Computenuorld, July 5,2004. Copyright © 2004 by Computerworld Inc., Framingham, MA 01701.All rights reserved.
1. IVR technology is advancing rapidly in terms of its abil-ity to simulate natural language conversations and ac-cept common phrases. Using the Internet, research thestate of the art in speech recognition. One good placeto start is www.scansoft.com.
2. Simply simulating natural conversations is but one ca-pability of speech software applications. Text-to-speech,voice verification, and speech-to-text are a few others.Break into small groups with your classmates, andbrainstorm how speech recognition systems can be usedin innovative and useful ways. Do you think we willeventually eliminate the need for humans in commontelephone interactions? Is this good or bad?