Telephonic communications – Audio message storage – retrieval – or synthesis – Multimedia system
Reexamination Certificate
2002-05-16
2004-08-03
Hoosain, Allan (Department: 2645)
Telephonic communications
Audio message storage, retrieval, or synthesis
Multimedia system
C379S088170, C379S088180, C379S265010, C379S265070, C379S265090, C379S266010, C379S266070
Reexamination Certificate
active
06771746
ABSTRACT:
This application relates generally to artificial intelligence using voice recognition and speech synthesis. More specifically, this application relates to a method and apparatus for automatic agents in an automatic call center using speech recognition and speech synthesis technology.
BACKGROUND
Call centers provide important solutions to customer contact problems in electronic commerce. An example of such a system is the Spectrum® Integrated Call Center System available from Rockwell Electronic Commerce, global provider of proven customer contact solutions for voice and Internet media. The Automatic Call Distributor (ACD) is a focal point of a customer contact center. The “calls” handled by an ACD can be any customer contact including but not limited to telephone calls, wireless calls, facsimile, e-mails, internet chat, and other internet like contacts. Thus, an ACD can accept customer voice and data contacts and distribute each to a properly skilled agent representative, and can be implemented on any structure using hardware, software or a combination of both in for example, an integrated system, a distributed system using multiple computers, or a software driven solution using a personal computer.
Call center administrators can set up a call flow or call flows for an ACD platform. Tools for managing call flows include scripts for routing calls. Such scripts can be programmed over a general-purpose computer included in an ACD platform.
Call routing allows a call center to accept each customer contact and distribute it to the proper agent representative. It is known to route calls using an icon-based Windows application tool. Using such a tool, a user can create a series of steps to route a customer contact efficiently to a desired agent. Scripts are used to direct calls to Agent Groups, Intelligent Announcements, ACD Mail, VRU ports and other call centers. Scripts can provide the call center with direct control call-by-call over interactions with the telephone network.
Existing technology requires that an individual agent handle each separate call. Under this existing technology, an individual agent must handle each call even if the call contact asks the same question and should receive the same responses as many other callers.
Known systems for handling calls in an automated call center require an extensive staff of agents, because an individual agent must participate in each call. Such individual call handling can be less than optimally efficient if a number of different call contacts tend to pose the same questions and should receive the same responses from agents. In addition to inefficiency, the quality of individual call handling can be affected by individual agent fatigue or by inadvertent errors or omissions by individual agents.
Current technology only allows for one agent per call. This requires an agent to handle each call even if the same questions or responses are given to or from one call contact as another call contact. If the agent is consistently responding to many callers in the same way it would be desirable to automate the process and only require an agent to be involved when something unexpected or unique happens during the call.
SUMMARY
The invention, together with the advantages thereof, may be understood by reference to the following description in conjunction with the accompanying figures, which illustrate some embodiments of the invention.
One mode of practicing the invention is a method of automatic call handling allowing agent optimization in an automatic call distribution system that comprises synthesizing speech by using a script as input and generating speech as output, connecting a call from or to a call contact, and speaking to the call contact using speech generated using the prepared script as input. The method further comprises identifying any need to use a live agent to respond to call contact, making an extemporaneous script using an agent speech recognizer to recognize live agent voice input and generating speech for transmission to the call contact. The agent speech recognizer can, in various embodiments, perform speech-to-text conversion, speaker dependent recognition, continuous-speech recognition, contiguous-word recognition, and isolated-words recognition. The speech synthesizer can also include additional steps to record different pronunciations from a professional voice, save the different pronunciations recorded from a professional voice, and use the different pronunciations recorded from a professional voice to synthesize speech.
A second mode of practicing the invention is an agent assist method of automatic call handling in an automatic call distribution that includes the steps of synthesizing speech by using a proposed script as input and generating as output speech that sounds like a live agent's natural voice, connecting a call with a call contact, and transmitting synthesized speech generated using the prepared script as input to the call contact. The method further comprises receiving call contact voice input, transmitting the call contact voice input to the live agent, and accepting direction from the live agent indicating a response to the call contact's voice input. Responses to the call contact's voice input can in some embodiments include providing a second prepared script and connecting the call to the live agent for the live agent to converse with the call contact. One embodiment of this mode may include steps to monitor background noise surrounding the live agent and then introduce background noise into the synthesized speech to mimic the live agent's environment.
Other modes of practicing the invention include a system with an automatic call distributor that has means to perform the steps of the method summarized above. Yet another mode of practicing the invention is a computer program product embedded in a tangible medium of expression that includes computer program code to perform the steps summarized above.
REFERENCES:
patent: 5325421 (1994-06-01), Hou et al.
patent: 5615296 (1997-03-01), Stanford et al.
patent: 5652789 (1997-07-01), Miner et al.
patent: 5757904 (1998-05-01), Anderson
patent: 5946386 (1999-08-01), Rogers et al.
patent: 6108610 (2000-08-01), Winn
patent: 6330079 (2001-12-01), Dugan et al.
patent: 6434230 (2002-08-01), Gabriel
patent: 6526395 (2003-02-01), Morris
Beck, III James L.
Shambaugh Craig
Hoosain Allan
Rockwell Electronic Commerce Technologies LLC
Welsh & Katz Ltd.
LandOfFree
Method and apparatus for agent optimization using speech... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for agent optimization using speech..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for agent optimization using speech... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3304965