Granted Patent B2

US 9,509,845 · App. 14/485,596 · Granted Nov 29, 2016

Abstract

Claims (56)

1. A system for pairing agents and callers within a call center environment, comprising:

two or more devices, each device comprising one of a conventional telephone handset, portable handset, and computer; and

a speech recognition server comprising a central processing unit, memory, input port to receive voice samples from one or more of the devices, and output port, and interconnected to each of the devices through one or more of Plain Old Telephone Service, cellular and satellite telephone service, and an internetwork, wherein the central processing unit is configured to execute modules comprising:

an analysis module to perform analyses on multiple voice samples of one or more candidate agents within a call center and to associate each of the voice samples from one such candidate agent with a different population for a common trait;

a call receipt module to receive a call from a caller into the call center and to receive a voice sample from the caller during the call;

a trait identification module to identify traits of the caller from the voice sample;

a comparison module to compare the traits of the caller to traits of one or more of the candidate agents within the call center;

a similarity module to determine a similarity between the caller and each candidate agent based a number of the traits shared by the caller and that candidate agent;

a selection module to select one of the candidate agents based on the similarity;

a call assignment module to assign callers to the selected candidate agent based on the different populations to which that candidate agent belongs; and

a call connection module to connect the caller with the selected candidate agent; a notice module to provide notice to the candidate agent upon assignment of each such caller, wherein the notice comprises identification of one of the populations associated with the candidate agent that is shared by one of the callers; and

a database to store the voice samples of the candidates agent and caller.

2. A system according to claim 1 , further comprising:

a measurement module to measure each trait of the caller and the candidate agents as one of a presence of the trait, an absence of the trait, along a continuum, and as a percentage.

3. A system according to claim 1 , further comprising:

a designation module to designate at least two of the traits of the caller as traits of interest;

an assignment module to assign a weight to each of the traits of interest; and

a similarity determination module to consider the weights of the traits when determining the similarity.

4. A system according to claim 1 , further comprising:

a sample module to obtain the voice sample from one such candidate agent, comprising:

an instruction module to provide instructions to the candidate agent to repeat one or more predetermined text passages into a telephone; and

a speech receipt module to receive verbal speech of the text passages from the candidate agent as the voice sample.

5. A system according to claim 1 , further comprising:

a trait determination module to determine the traits of the caller and each candidate agent by determining voice characteristics from the voice sample comprising at least one of a pitch, range of resonance frequencies of vowels, an average resonance frequency of predetermined vowels, spectral tilt, breathiness, creakiness, nasality, speed of words, pauses, and intonation for each voice sample and by identifying the traits associated with the determined voice characteristics.

6. A system according to claim 1 , further comprising:

a trait determination module to determine a suitability of one such candidate agent to assist the caller based on at least one of comprehensibility, perceived appeal to one or more audience types, competence at various tasks and skills, knowledge, and access to relevant information.

7. A system according to claim 1 , wherein the number of shared traits can be determined based on one of an absolute number, a threshold, and as a highest number of shared traits.

8. A method for pairing agents and callers within a call center environment, comprising steps of:

monitoring two or more devices, each device comprising one of a conventional telephone handset, portable handset, and computer;

maintaining a speech recognition server comprising a central processing unit, memory, input port to receive voice samples from one or more of the devices, and output port, and that is interconnected to each of the devices through one or more of Plain Old Telephone Service, cellular and satellite telephone service, and an internetwork;

performing via the speech recognition server multiple analyses on the voice samples of one or more candidate agents; and

associating via the speech recognition server each of the voice samples from one such candidate agent with a different population for a common trait;

receiving a call from a caller into a call center via at least one of the conventional telephone handsets, portable handsets, and computers;

receiving the voice sample from the caller during the call and identifying traits of the caller from the voice sample via the speech recognition server;

comparing via the speech recognition server the traits of the caller to traits of one or more of the candidate agents within the call center;

determining via the speech recognition server a similarity between the caller and each candidate agent based a number of the traits shared by the caller and that candidate agent;

selecting one of the candidate agents based on the similarity by the speech recognition server;

assigning the caller to the selected candidate agent based on the different populations to which that candidate agent belongs; and

connecting the caller with the selected candidate agent; and providing notice to the candidate agent upon assignment of each such caller, wherein the notice comprises identification of one of the populations associated with the candidate agent that is shared by one of the callers.

9. A method according to claim 8 , further comprising:

measuring each trait for the caller and the candidate agents as one of a presence of that trait, absence of that trait, along a continuum, and as a percentage.

10. A method according to claim 8 , further comprising:

designating at least two of the traits of the caller as traits of interest;

assigning a weight to each of the traits of interest; and

considering the weights of the traits when determining the similarity.

11. A method according to claim 8 , further comprising:

obtaining the voice sample from one such candidate agent, comprising:

providing instructions to the candidate agent to repeat one or more predetermined text passages into a telephone; and

receiving verbal speech of the text passages from the candidate agent as the voice sample.

12. A method according to claim 8 , further comprising:

determining the traits of the caller and each candidate agent, comprising:

analyzing the voice samples by determining voice characteristics from the voice sample comprising at least one of a pitch, range of resonance frequencies of vowels, an average resonance frequency of predetermined vowels, spectral tilt, breathiness, creakiness, nasality, speed of words, pauses, and intonation for each voice sample; and

identifying the traits associated with the determined voice characteristics.

13. A method according to claim 8 , further comprising:

determining a suitability of one such candidate agent to assist the caller based on at least one of comprehensibility, perceived appeal to one or more audience types, competence at various tasks and skills, knowledge, and access to relevant information.

14. A method according to claim 8 , wherein the number of shared traits can be determined based on one of an absolute number, a threshold, and as a highest number of shared traits.

Assignments (15)